A response to a post by Mr. Yourdon

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread
I wrote the following response to Mr. Yourdon on another thread. I thought the issues were well worth discussing.

"Mr. Yourdon,

Acknowledging that information systems is not my field, let me suggest the following points:

1. Software remediation is not software development or even software engineering. Performance data on the development of a new operating system is not applicable to remediating a specific data element in an existing program. It is like comparing building a car to repairing a carburator.

2. The process of remediation has been increasingly automated thanks to software tools developed in the private sector. Finding Y2K- related problems has become easier and faster (due to market forces). And there is no reason to believe the process will not accelerate as the year continues. Oh, and Y2K programmers are not making NBA salaries yet.

3. Fixing noncompliant systems is very high priority for most firms. As you know, many software projects are doomed from the start. They fade because they lack a clear vision, a well conceived functional analysis, appropriate staffing and resources, etc. It is incorrect to compare an organization's Y2K efforts to "just another software project." The level of attention is closer to a response to a virus infection or a systems "hack."

4. Not all firms rely on computers. Not all computers are "mission critical." Not all programs contain date sensitive data. Not all Y2K noncompliant programs will fail. We do not have data on the percentages, but it is misleading to suggest a single date field error will incapacitate an entire IS system.

Many small businesses use computers in a support role. It is possible for most firms to function, even when the compters are "down," although less efficently.

5. How much of government is "mission critical?" If the National Endowment for the Arts (NEA) has a systems failure, can we expect angry mimes conducting street demonstrations? Most federal and state agencies could close tomorrow with very little impact aside from the dislocated workers. In fact, it might actually improve the business environment.

6. The U.S. financial system is a widely acknowledged leader in Y2K remediation. Even one-year CDs now mature in 2000. Most authorities in the the critical infrastructure systems--telecommunications, utilities and finance--have positive reports about Y2K remediation efforts. Please, Mr. Yourdon, I hope you are not suggesting, like some readers, a massive conspiracy of silence.

7. With your extensive experience, you know that systems lurch along with patches, work-arounds, quick fixes, manual adjustments, etc. Yes, some firms will fail because they have a complete systems failure, but it takes time to fail. If the network goes down, Mr. Yourdon, the employees do not begin cleaning the personal items out of their desk. Companies will exhaust cash reserves, hire additional staff; they will do everything possible to keep the business moving forward. And, as always, the IS staff will work long hours.

7. The rest of the world's problems (and they are quite extensive) have not slowed our domestic economy. Honestly, I do not think Kenya is a major trading partner... nor do I think many lesser-developed countries (LDCs) are as technology dependent as the advanced economies. Of course, if there is a global depression it will impact the United States... but read some of the posts here. Do you think we are headed for martial law? Do you think the Republic is so fragile that a computer systems problem will result in anarchy?

8. There is no "the code." There are countless hardware platforms and software packages. There is no "broken." Some specific elements must be remediated to ensure Y2K compliance. In this case, Mr. Yourdon, we are looking for a "good enough" solution.

9. Individual firms will fail. If we allow capitalism to work, they always will. Failure, within reason, is a good thing. In fact, we probably do not let enough firms perish. Our government "saves" some businesses that really ought to die. Our government also has many agencies of dubious value. If they do not function, it may prove that we can function as a society quite well without them.

10. You argue that Y2K is like every other software project we have faced during the last 30 years. It started late and will finish late. You also argue that Y2K is unlike anything we have faced before because it is "systemic." If it is "unlike" anything we have faced, isn't it reasaonble to expect a unique response (like the considerable increase in your public profile).

If you were arguing that one "ultra virus" could disable massive numbers of computers, corrupt data and cause serious economic disruptions... I'd agree. In fact, I think we'll see cyber terrorism in our lifetimes. If you think we ought to prepare for such contingencies, again, I agree.

To use a baseball analogy, Y2K is a slow pitch right down the middle of the plate. Everyone in the stadium knows it... and you are right, some firms will still strike out. The pitch I worry about is the wicked split-finger fastball that no one expects.

Thank you for your consideration and I look forward to further conversations.

Regards,

Mr. Decker"



-- Mr. Decker (kcdecker@worldnet.att.net), April 07, 1999

Answers

Mr. Decker,

Thanks for your interesting and thoughtful reply to my earlier message. I've inserted my responses in between your comments...

1. Software remediation is not software development or even software engineering. Performance data on the development of a new operating system is not applicable to remediating a specific data element in an existing program. It is like comparing building a car to repairing a carburator.

I commented on this at length in the essay I referred to last night, Y2K Software Projects: Deja Vu All Over Again. I see also that someone else added a note to the thread, suggesting that "late starts" was not the biggest project risk, but that lack of clear requirements was the biggest problem.   This often comes up in such discussions, because people (including computer professionals) sometimes argue that Y2K projects are simpler because the requirements are clear. But consider the following points:

2. The process of remediation has been increasingly automated thanks to software tools developed in the private sector. Finding Y2K- related problems has become easier and faster (due to market forces). And there is no reason to believe the process will not accelerate as the year continues. Oh, and Y2K programmers are not making NBA salaries yet.

Deja vu all over again: we have CASE tools and RAD tools and reusable components marketed on the Internet, and a variety of other such things in the "traditional" software development arena. As for Y2K: yes, there are automated solutions for COBOL and C and a few other popular languages. But we're coping with 500 different programming languages, and the "market forces" have intelligently concluded that there's no point developing automated tools for obscure languages used by only a fraction of the IT industry. Why should the process accelerate as the year continues? We're supposed to be testing now, and while there are automated testing tools, that too is deja vu all over again. As for Y2K programmers making NBA salaries ... not sure what you meant by that.

3. Fixing noncompliant systems is very high priority for most firms. As you know, many software projects are doomed from the start. They fade because they lack a clear vision, a well conceived functional analysis, appropriate staffing and resources, etc. It is incorrect to compare an organization's Y2K efforts to "just another software project." The level of attention is closer to a response to a virus infection or a systems "hack."

If it's a very high priority, then why have most companies devoted only 25% of the IT resources to Y2K? Why do most public-sector government agencies (especially at the state and local level) find that their Y2K budget requests are repeatedly cut in half, or rejected completely?   The most hilarious example of this that popped up on the Internet a couple days ago is the city of Beijing (I realize that this is a minor insignificant city that has nothing to do with the U.S., but it's amusing nonetheless) that is trying to accomplish its Y2K remediation for a population of nearly 10 million people with a total budget of $500,000. In general: fixing noncompliant systems is a high priority only in those companies where middle management and the techies understand that it's a high priority, and where senior management has actively, enthusiastically, and repeatedly articulated that priority. Obviously, there are some companies that have done so; there are a lot more companies that view Y2K as an annoying nuisance that they have to go through; and there are some companies that have done nothing more than send an email message to the programmers that says "take a look at your programs, and if you see any Y2K problems, fix them up ... in your spare time." I'm not joking about this last point; I've visited companies with that attitude. It's a small percentage of the overall industry, but I argue that we're dealing with a bell-curve phenomenon here; it's deja vu all over again.

4. Not all firms rely on computers. Not all computers are "mission critical." Not all programs contain date sensitive data. Not all Y2K noncompliant programs will fail. We do not have data on the percentages, but it is misleading to suggest a single date field error will incapacitate an entire IS system.

Yeah, the local pizza shop doesn't rely on computers -- but if the lights are out, the phones are dead, and the banks are closed, the pizza shop is still going to be affected by Y2K. And aside from the tiny mom-and-pop businesses, I think it's fair to say that most large companies do rely on computers. Even the small-to-medium companies are pretty dependent today on LC's and LANs for order entry, invoicing, accounting, and other traditional business functions. Obviously, not all computers are "mission-critical" -- but all of the discussions about Y2K progress and compliance (e.g., the recent statement by the federal covernment that it achieved 92% compliance by March 31, 1999) are focused only on the mission-critical systems.

As for suggestions that a single date field error will incapacitate an entire IS system -- I'm not sure if I ever said such a thing, but the appropriate answer is (a) it depends, and (b) on large IS applications, chances are that we're going to be experiencing much more than just one such error. I'm not sure if you're aware of it, but we're now beginning to get some specific data from the IV&V reviews, conducted by such firms as Cap Gemini, Reasoning Systems, and MatriDigm. These are reviews of supposedly-remediated, and supposedly-tested code, in quantities approaching 50 million lines of code. The IV&V reviews have typically found between 450 and 900 date-related bugs per million lines of code.   Of these, roughly 30% were classified as "serious" errors, in contrast to minor formatting problems or slightly confusing displays or reports. What's significant about this is that the IV&V efforts have taken place with clients that area ready for them to do so -- i.e., they got started early enough that they have now finished their remediation, and finished all of their own internal testing; one has to assume that they believe they've done a good job, and that they've done enough testing to be ready to put the systems back into production. The organizations that got started late won't be in such a position -- another one of the "deja vu" aspects of Y2K is that testing always takes place at the end of the project, and if you're behind schedule, you simply do less testing. Thus, a lot of the supposedly-remediated Y2K code that is declared "finished" during the second half of the year will be code that was not tested as thoroughly as it should have been; thus the 450-900 bugs/million lines of code is probably an optimistic, if not best-case, estimate of the quality we can expect.

Many small businesses use computers in a support role. It is possible for most firms to function, even when the computers are "down," although less efficently.

If the small business has one PC, I would agree with you. If it has 5-10 such machines, connected on a LAN, I think you'll find that the PC is a critical component of their daily operation. One of my clients is a $6 million/year publishing business that has approx 40 employees and 40 computers. If the network goes down (i.e., their internal LAN), they can continue to do some stand-alone work, but their overall productivity suffers. If the Internet goes down, they suffer more; if their billing and order-entry systems go down, they are in serious trouble.

5. How much of government is "mission critical?" If the National Endowment for the Arts (NEA) has a systems failure, can we expect angry mimes conducting street demonstrations? Most federal and state agencies could close tomorrow with very little impact aside from the dislocated workers. In fact, it might actually improve the business environment.

I assume that DoD is essential. So is HHS, IRS, SSA, DOT, Treasury, and a few others. If the Department of Education stops sending money to schools and colleges, I suspect you'll hear about it pretty quickly. I agree that we can probably survive without NEA and AID and a few other operations...

6. The U.S. financial system is a widely acknowledged leader in Y2K remediation. Even one-year CDs now mature in 2000. Most authorities in the the critical infrastructure systems-- telecommunications, utilities and finance--have positive reports about Y2K remediation efforts. Please, Mr. Yourdon, I hope you are not suggesting, like some readers, a massive conspiracy of silence.

"Conspiracy of silence" is far too emotional a term. What I would suggest -- from the perspective of having run a couple small companies, and also from having served on Boards of Directors -- is that senior management of private-sector organizations are reluctant, if not totally unwilling, to stand up and broadcast negative information about their companies.   This is especially true of publicly-traded companies, and is even more true with the threat of Y2K-related litigation. I agree that virtually all major private-sector organizations are working on Y2K, and I agree that there is good news to be reported from such industries as telecommunications, utilities, and finance. But I would argue that senior managers of these organizations are prone to report only the good news, for fear that their stock might plummet otherwise, or that they might be subject to class-action lawsuits. Indeed, if I wanted to suggest something along the lines of a "conspiracy," it would have to do with the lack of any third-party, independent audits or reviews of the alleged Y2K compliance of these organizations. In the case of the banking industry, for example, the Fed specifically and explicitly prohibits a bank from revealing the results of the Y2K inspections that have been taking place. Here's the link to Fed regulation, and here's the relevant language

"Information from Year 2000 assessments are governed by the same rules of confidentiality that apply to FDIC examinations for safety and soundness, compliance, information systems, and trust activities. Under Part 309 of the FDICs rules and regulations, disclosure of reports of examination, or any information contained in them, is strictly prohibited. Accordingly, institutions may not disclose results from Year 2000 assessments just as they may not disclose other types of examination information. "

"Moreover, disclosure of such information to third parties such as financial ratings firms or fidelity bond carriers is likewise prohibited. Requests from such entities are not authorized by the FDIC or any other banking regulator. In light of the blanket prohibition on disclosing ratings, compilations of Year 2000 ratings by such firms are necessarily incomplete and unreliable."

This puts the average citizen in a position where he either has to (a) accept all of the PR reports from the government and banks at face value, or (b) rely on his own experience and access to information. The average citizen probablyh doesn't have access to any other independent source of information, but in today's environment, he tends to be rather cynical and dubious about any official government statements; perhaps that's why we continue to see the public-opinion polls indicating that roughly 15% of the public plan to withdraw most or all of their money.   In my case, I have 35 years of IT experience with large organizations of all kinds, including banks. My experience is that banks are aggressive users of IT technology, but they're not rocket scientists -- they've managed to screw up large, complex projects just like everyone else. When I read the 10-Q SEC statements indicating that several banks are spending $300-500 million on Y2K, and employing armies of 1,000 programmers to fix their computer systems, I can't help asking myself, "When is the last time I saw that bank, or any bank, deliver a thousand-person, $300 million project on time?"

7. With your extensive experience, you know that systems lurch along with patches, work-arounds, quick fixes, manual adjustments, etc. Yes, some firms will fail because they have a complete systems failure, but it takes time to fail. If the network goes down, Mr. Yourdon, the employees do not begin cleaning the personal items out of their desk. Companies will exhaust cash reserves, hire additional staff; they will do everything possible to keep the business moving forward. And, as always, the IS staff will work long hours.

Yes, they do have patches, workarounds, and daily crashes, aborts, ABENDs, etc. Thus, as I've argued in several of my articles and essays, the key questions are: (a) how many more failures are they likely to experience because of Y2K, and even more important, (b) how long does it take them to repair the failures and resume more-or-less normal operations? Both MTBF (mean time between failures) and MTTR (mean time to repair) are going to be critical factors in determining whether Y2K is a minor nuisance or a major catastrophe. A few of the more proactive firms (including one financial institution that I know) have been attempting to gather some preliminary data about this, based on their Y99 failures that occurred on or about 1 Jan 1999; but most companies have no data to help them plan for this. But I certainly agree with your point that companies will not shut their doors right away; they'll invoke whatever contingency plans they've developed, they'll do their best to employ manual workarounds, etc. And thus, one might make a plausible case that we'll avoid TEOTWAWKI by virtue of these fallback procedures; but don't you think we'll also see the consequences in terms of lost productivity, lower revenues, higher error and re-work rates, etc.?

7. The rest of the world's problems (and they are quite extensive) have not slowed our domestic economy. Honestly, I do not think Kenya is a major trading partner... nor do I think many lesser-developed countries (LDCs) are as technology dependent as the advanced economies. Of course, if there is a global depression it will impact the United States... but read some of the posts here. Do you think we are headed for martial law? Do you think the Republic is so fragile that a computer systems problem will result in anarchy?

I suspect that your knowledge of economics will overwhelm any comments I can make in this area ... but my impression was that many of our national leaders have been concerned about the impact of problems in Asia, South America, and Russia on our economy. I vaguely recall reading that exports of high-tech companies to Asia have fallen, because of financial problems in Korea, etc. I also vaguely recall reading that the recent devaluation of the Brazilian currency posed an economic threat to our economy, in terms of the impact on imports and exports. Obviously, Kenya alone is not a major trading factor; I was using it as an example of the total lack of Y2K preparedness that one sees in most of the devfeloping nations. As for the question of martial law: I doubt that it will necessary or appropriate in the little town where I live, but it wouldn't surprise me to see such actions in various large cities, if they are impacted by a combination of infrastructure failures and civil unrest. All it took was a handful of cops bopping poor Rodney King over the head to cause a significant portion of Los Angeles to be burned down a few years ago ... and I think that involved curfews and National Guard troops patrolling the streets, if not outright martial law.

As for the issue of lesser-developed countries (LDCs) being technologically dependent ... this turns out to be a somewhat more complex issue than one first imagines. Yes, the "good news" is that they have less computer technology; but the "bad news" is that their technology is older, and is often pirated software they can't fix through normal channels. The good news is that most of their computer technology is centralized (because the governments control the banks, the phone companies, and other major infrastructure elements) ... but the bad news is that the governments of the LDCs don't have the money, or the talent, or the sense of urgency to fix their computers that one would expect in the private sector.

As for the question of whether the Republic is so fragile that a computer systems problem will result in anarchy ... well, remember that you were the one who started the discussion last night by saying that we were dealing with a "business problem." I would suggest that we've got a political problem, sitting on top of a business problem, which sits on top of a computer technology problem. If it turns out that the computer technology problem is nothing more than a "bump in the road," then the two layers above won't matter. If it's a moderately serious problem, it will impact the business community and perhaps lead to a recession, but it doesn't necessarily affect the political structure of the country. But if it turns out that Y2K causes major infrastructure failures in several large cities, and if it's compounded by global problems and a severe economic recession, then it seems to me that the political structure will be affected. Look what the Great Depression did to our political structure in the 1930s; look what it did to Europe; look what it did to the entire global political environment.

One interesting aspect of this is the relative lack of national leadership, the relative silence from state leaders, and the predominance of local grass-roots activities for Y2K preparedness. I attended a local Y2K community meeting in my town about a month ago, in which an official from the state-government emergency-services department said to an audience of about 300 people, "If Y2K turns out to be a serious problem, you're on your own. You're welcome to give us a call at the state capitol, and we'll help you if we can; but the chance are that if you're having problems with utilities and water and telecommunications, then we're having bigger problems. You're on your own. You can ask us to call Washington for federal assistance -- but if you're having trouble, and we're having trouble in the state capitol, then there's a pretty good chance that they're having far worse problems in the big cities. You're on your own. You're on your own."

Under the circumstances, I think it was a perfectly reasonable statement for this official to make. But I think it could have significant political ramifications, especially when you consider that there are now some 300 local Y2K community groups listed in the Cassandra Project's index. If Y2K does turn out to be a serious problem, and if communities across the land realize that not only are they on their own, but they were given no advance guidance and advice from the federal government (aside from the simplistic advice to treat Y2K like a winter snowstorm), then the grass-roots Y2K community groups could be the foundation for a populist movement that sees the 2000 elections as an opportunity to achieve a massive transfer of political power from the federal and state officials to the local level.

8. There is no "the code." There are countless hardware platforms and software packages. There is no "broken." Some specific elements must be remediated to ensure Y2K compliance. In this case, Mr. Yourdon, we are looking for a "good enough" solution.

Okay, you're right: we don't have a homogeneous environment, but rather an incredibly heterogeneous, eclectic environment of hardware and software. But I would argue that that complicates the problem, because we don't have any possibility of a single "silver bullet" solution. Nonetheless, I do agree with you that we're looking for a "good enough" solution ...ironically, that will come back to haunt us (e.g., in the case of the "windowing" solutions to Y2K) twenty or thirty years from now, when our "good enough" solutions fail once again. ...

9. Individual firms will fail. If we allow capitalism to work, they always will. Failure, within reason, is a good thing. In fact, we probably do not let enough firms perish. Our government "saves" some businesses that really ought to die. Our government also has many agencies of dubious value. If they do not function, it may prove that we can function as a society quite well without them.

I completely agree with you about this, and I would be one of the last people to suggest that the federal government shoujld "save" any businesses. Unless Y2K turns out to be so bad that it knocks us back to the Stone Age (which I don't believe will happen), then the economy will eventually recover, and the various social institutions will eventually recover. But "eventually" can be a long time; and while I may well be proven wrong, I don't see it happening within one calendar quarter in the first three months of 2000.

10. You argue that Y2K is like every other software project we have faced during the last 30 years. It started late and will finish late. You also argue that Y2K is unlike anything we have faced before because it is "systemic." If it is "unlike" anything we have faced, isn't it reasaonble to expect a unique response (like the considerable increase in your public profile).

I'm not sure how to respond to this; my public profile is irrelevant to the outcome of Y2K. If indeed our federal government had viewed Y2K as a unique and potentially devastating problem, then it might have been able to respond in a unique fashion. One of the interesting points made by Dr. Paula Gordon during her presentations when I visited Washington last week is that the Federal government has created an organizational response (in the form of the Y2K Conversion Council headed up by Mr. Koskinen) that has the characteristics of an organization dealing with a "routine" situation; Mr. Koskinen's charter is to coordinate, collaborate, and raise awareness -- but with a staff of only 11 people, he can't actually do anything. If the government had created an organization similar to NASA, or the Federal Energy Emergency program, or the various organizations that have been set up to cope with the drug problem, then one could imagine that they really did believe that it was unique. (By the way, I'm not trying to suggest that the drug programs or other emergency- oriented organizational structures were necessarily successful...)

If you were arguing that one "ultra virus" could disable massive numbers of computers, corrupt data and cause serious economic disruptions... I'd agree. In fact, I think we'll see cyber-terrorism in our lifetimes. If you think we ought to prepare for such contingencies, again, I agree.

As you know, some viruses (like Michelangelo) turned out to be a hoax; others, like the recent Melissa virus, turned out to be harmful -- but were capable of being stopped before they spread worldwide. The nature of Y2K is such that it is world-wide, and (as you mentioned above), it will manifest itself in an enormous variety of hardware/software environments. That's one of the reasons I'm so worried about it.

To use a baseball analogy, Y2K is a slow pitch right down the middle of the plate. Everyone in the stadium knows it... and you are right, some firms will still strike out. The pitch I worry about is the wicked split- finger fastball that no one expects.

Take a look at the essay I wrote last year, entitled "What if Y2K Was A Baseball Game?". In a baseball game, we know what the pinch-hitter's batting average is before we make the decision to send him to the plate in the 9th inning, with the bases loaded. We have lots of statistics that help us make an intelligent prediction of the outcome, in terms of batting and pitching and fielding. We would be extremely annoyed if we asked the pinch-hitter for his batting average, in order to evaluate our chances of success, only to have him say to us, "I don't keep no stinkin' batting average! And if I did, it wouldn't matter anyway, because things are different this time! You just gotta trust me!"

Ed

-- Ed Yourdon (ed@yourdon.com), April 07, 1999.

Lurky, Yeah, there are advantages to being the sysop -- I had typed my message into Netscape Composer, and thought the HTML formatting would be okay. But when I pasted it into this tiny little message window, strange things happened....

Someone, eh? I thought I was making a point-by-point response to Mr. Decker's comments.

Mr. Decker, No need to respond if you're busy -- I certainly sympathize! And I suspect that it could be an endless discussion, since we may be looking at the problem from such different perspectives, and with such different assumptions. When the history of Y2K gets written, I suspect that we'll find it's a multi-faceted phenomenon and that it can't be described or explained in terms of a single-dimension "truth". Your points about the vitality of the free-enterprise system certainly have a lot of merit, but I hope that several of the postings on this thread have provided a coherent statement of the concerns of some (but not all) of the computer professionals...

Ed

-- Ed Yourdon (ed@yourdon.com), April 07, 1999.

I've stated your first point on this forum in the past. I've been in software development for twenty years and agree that remediation is not the same. The big problem in development is not that projects start late (as Ed pointed out on the other thread) but that requirements are not clearly defined from the beginning and that the user constantly changes them. I don't think I have ever seen that statistic that projects start late from Dr. Rubin or Casper Jones or the Garnter Group. Project Management states that planning the project up front is most important to the success of the project and many just jump right into the coding phase. Well none of this applies to Y2K. In Y2K the requirements are clearly defined, the design is set, the solutions are limited. The most important task in Y2K lies in analysis; find all the infections. And as you point out, tools (available because of our free enterprise) make that job easier than ten years ago when some Y2K projects started.

-- Maria (anon@ymous.com), April 07, 1999.
Some interesting observations but here is my counterpoint on it.

The parts in italics are quotes from the original message.

1. Software remediation is not software development or even software engineering.

Granted, however I can imagine in quite a few instances it would be better to implement a set of new software to get the job done than remediate an old set. The company may want extra features in their software and now it gives them an excuse to write new software from scratch that has all the things that they have wanted all this time. They can either remediate now and then do the new software later or save money and just do the new software now.

So there will be some companies producing new software, others who would be remediating and yet others who aren't even going to bother.

2. The process of remediation has been increasingly automated thanks to software tools developed in the private sector.

This is certainly no guarantee that such a magical piece of software could find every single occurance and fix it properly. Someone would still need to go through manually just to make sure. I can't imagine any person totally relying 100 percent on any software.

3. It is incorrect to compare an organization's Y2K efforts to "just another software project." The level of attention is closer to a response to a virus infection or a systems "hack."

As recently shown the Melissa macro virus was in major headlines for several days running, has Y2K received this level of attention? Within 24 hours of me hearing about the macro virus I had the computers at where I work ninety-nine percent protected against it.

My employer is more worried about that virus or any virus, more worried about system hacks (Every time something goes wrong with our computers he thinks the system has been hacked.) and worries zilch about Y2K. He laughs at people and proclaims to all and sundry that we are Y2K compliant, he hasn't checked any of our systems yet and he definately has not asked me to check either.

I got more queries in a few days about the Melissa macro virus than I have about Y2K in 6 months.

4. Not all firms rely on computers

Most do however, print letters, accounting, names & addresses, meetings, graphic design, computer shops, internet service providers, etc.

If you couldn't print letters, granted you would write them out. But this is time consuming and harder to rectify mistakes and check your spelling. :-) The inconvenience alone of having to do most things by hand would delay things significantly.

Not all programs contain date sensitive data.

You still have to check every single program, just in case. There was one case of a failure (sorry I can't provide a reference at the moment but I do remember it being listed amongst Rob Michael's list of failures) where the company fixed over 100 pieces of software and ignored this one little program that they thought was not important.

It crashed the system, totally.

Personally I'm an idiot, I must be if I keep making changes to a piece of software and never checking to see if it would affect any others that it may interact with. I've lost count of how many problems I've experienced because I've overlooked something or I thought that it wasn't important.

but it is misleading to suggest a single date field error will incapacitate an entire IS system.

No it isn't, because a single little itty bitty date field can. On the Australian 60 minutes it had a segment about Y2K, in one part they gave an example of a bank that discovered that on the 1st January 2000 every single one of their customers accounts would have shutdown and been closed. Do you think the bank would have been able to function afterwards? I doubt it, customer confidence would have plummeted to zero, and the time to fix it would be mind-boggling. Goodbye bank.

5. Most federal and state agencies could close tomorrow with very little impact aside from the dislocated workers.

How many other people/companies interact with those agencies either as a support role or financially. Those dislocated workers will need to find new jobs, so suddenly your jobless rate is higher and you are getting even more people applying for those rare elusive jobs because you have to compete with twice or more people as before to get it.

An example, a city has 1 million people, out of those 10% are unemployed and looking for work. They receive welfare benefits, food stamps, handouts, etc to keep going. Suddenly in one week 10% of businesses close down and 10% of employees get laid off. Now suddenly you have 20% unemployed requiring more welfare, more food stamps and more handouts. But wait there is more, because of people having a drastic reduction in income they cannot afford to buy as much or pay bills as much anymore. Now suddenly 10% less is spent at the local grocery store, well that store doesn't need as many employees anymore to keep functioning and it is chewing into profits and costs, so sorry we have to let you go.

People default on rent, mortgages, loans and other payments. Most will start selling stuff, to who? You don't have as many people purchasing stuff any more and the new 10% is trying to sell stuff as well to keep living. Suddenly your $400 television is competing with the hundreds of others in the local newspaper.

This is called the domino effect, we've only ever had localized minor interruptions. The rest of society adjusts to it, but if 10% of people lost their jobs over the next week then that would have repercussions for months. Just look at Russia.

6. telecommunications, utilities and finance--have positive reports about Y2K remediation efforts.

But it isn't totally 100% fixed, yes it is good news. But it would be better news if it was all fixed.

That is all I have time for tonight, it is getting late (middle of the night here) and I must be abed as I have to go to work again tomorrow.

Regards, Simon



-- Simon Richards (simon@wair.com.au), April 07, 1999.

Somehow I missed this thread, and started another with the following:

Mr. Decker,

Your statement, "Acknowledging that information systems is not my field, let me suggest the following points:", was most appropriate as it excuses your erroneous conclusions in several cases.

"1. Software remediation is not software development or even software engineering. Performance data on the development of a new operating system is not applicable to remediating a specific data element in an existing program. It is like comparing building a car to repairing a carburator."

Your comparison is inappropriate. "Remediation" is a Y2K generated buzzword that refers to the process of making software code designed to work only in the 20th century, work after that century ends. As such, the comparison, to be accurate should be that of building a car to re-designing a part of that car and re-building it, without the old part and incorporating the re-designed one.

"2. The process of remediation has been increasingly automated thanks to software tools developed in the private sector. Finding Y2K- related problems has become easier and faster (due to market forces). And there is no reason to believe the process will not accelerate as the year continues. Oh, and Y2K programmers are not making NBA salaries yet."

Although some automated software tools have been developed for some applications, and this process may accelerate, they have not and will not replace the human programmers to the extent that would be required to locate all the faulty code. In addition, such tools seldom do more than locate date related code. That code must still be re-written by a human programmer. As for the failure of programmers' salaries to skyrocket as predicted, you may look to the greed of the corporate stockholders as impetus (after all, greed by whatever name you choose to call it, drives capitalism) and the foolhardy dispatch of such projects to India by "horn-haired" managers and hastily altered immigration quotas by the federal government as the explanation.

"3. Fixing noncompliant systems is very high priority for most firms. As you know, many software projects are doomed from the start. They fade because they lack a clear vision, a well conceived functional analysis, appropriate staffing and resources, etc. It is incorrect to compare an organization's Y2K efforts to "just another software project." The level of attention is closer to a response to a virus infection or a systems 'hack.'"

You are undoubtedly correct here, but the conclusion that such attention and priority will affect success is not supported. The crew of a naval destroyer that watches a spread of torpedoes unerringly seeking them as a target will certainly give the utmost attention to the "problem" and will no doubt respond in the extreme, but such will not ensure success in their case either.

"4. Not all firms rely on computers. Not all computers are "mission critical." Not all programs contain date sensitive data. Not all Y2K noncompliant programs will fail. We do not have data on the percentages, but it is misleading to suggest a single date field error will incapacitate an entire IS system.

Many small businesses use computers in a support role. It is possible for most firms to function, even when the compters are "down," although less efficently."

All firms do rely in computers! I suspect that you meant to say that not all firms rely on their own computers. That simply makes the problem worse from that firm's perspective since they have less control over any attempt to solve the problem.

It is quite correct that not all computers are "mission critical". Loss of such systems will (presumably) simply degrade the effectiveness of the operation of which they are a part.

It is also correct that not all programs contain date sensitive data. To be totally accurate, most programs do not contain data, they contain "code", but your meaning is clear and those programs that do not use dates will not fail and are thus "Y2K Irrelevant".

By definition however, a Y2K noncompliant program must fail. If it didn't, it would be "Y2K Compliant. Did you perhaps mean that such failures would not matter? That is one of the results that will occur.

And, while it is misleading to suggest a single date field error will incapacitate an entire IS system, it is not misleading to suggest a single date field error can or may incapacitate an entire IS system.

"5. How much of government is "mission critical?" If the National Endowment for the Arts (NEA) has a systems failure, can we expect angry mimes conducting street demonstrations? Most federal and state agencies could close tomorrow with very little impact aside from the dislocated workers. In fact, it might actually improve the business environment."

This is an excellent point of consideration and is demonstrated amply by the federal government's own classification of such a small percentage of its systems as "mission critical". Certainly "angry mimes" are not to be expected, nor are angry taxpayers when the IRS fails. Nevertheless, those few "mission critical" systems (IRS, Social Security, FMS, etc.) must function in order to prevent the fall of the government. The remainder should have been the subject when those of us who visit the ballot box have done so in the past. Most immediately, they seem irrelevant to Y2K.

"6. The U.S. financial system is a widely acknowledged leader in Y2K remediation. Even one-year CDs now mature in 2000. Most authorities in the the critical infrastructure systems--telecommunications, utilities and finance--have positive reports about Y2K remediation efforts. Please, Mr. Yourdon, I hope you are not suggesting, like some readers, a massive conspiracy of silence."

Conspiracy, by definition, requires interaction between the participants and an agreed upon scheme, to some unlawful end. I submit that such is not only highly unlikely (as you suggest), but unnecessary in order to observe the results that we do. What we are seeing in not conspiratorial behavior, it is normal, everyday, business-as-usual behavior.

"7. With your extensive experience, you know that systems lurch along with patches, work-arounds, quick fixes, manual adjustments, etc. Yes, some firms will fail because they have a complete systems failure, but it takes time to fail. If the network goes down, Mr. Yourdon, the employees do not begin cleaning the personal items out of their desk. Companies will exhaust cash reserves, hire additional staff; they will do everything possible to keep the business moving forward. And, as always, the IS staff will work long hours."

I agree with you here, but the description you provide is hardly cause for celebration. As has been discussed, that which has not been solved in a long period of time is unlikely to be solved in a short period of time. There will be exceptions (Oh! So that's what's wrong!) but in the main, the problem may be expected to get much worse before it gets better. In the meantime, events may well occur which will be greatly ameliorated by some simple forethought and preparation, at all levels.

"7. The rest of the world's problems (and they are quite extensive) have not slowed our domestic economy. Honestly, I do not think Kenya is a major trading partner... nor do I think many lesser-developed countries (LDCs) are as technology dependent as the advanced economies. Of course, if there is a global depression it will impact the United States... but read some of the posts here. Do you think we are headed for martial law? Do you think the Republic is so fragile that a computer systems problem will result in anarchy?"

I basically agree again, but you ask two questions. Martial law is simply not a viable option, for logistical reasons if no other. As for the fragility of the Republic, I do indeed perceive it to be that fragile. Whether or not anarchy results, the results will not be pretty and again, simple forethought and preparation at all levels is appropriate.

"8. There is no "the code." There are countless hardware platforms and software packages. There is no "broken." Some specific elements must be remediated to ensure Y2K compliance. In this case, Mr. Yourdon, we are looking for a "good enough" solution."

Here is one of the places where your ignorance of computer system internal workings has gotten you in trouble. There is indeed "the code", but you are right in that it is not "broken". "The code" consists of the simple assumption that whatever system is at issue is operating in the 20th century. As such, the code is not "broken"; it is doing exactly what it was designed to do. The problem is that soon it will not be the 20th century anymore and we will want "the code" to do something else. Your "good enough" solution is just what everyone has been madly scrambling to achieve and will largely fail at achieving. It has been apparent from the start (at least since 1996 or so) that a complete redesign of all code and restructuring of all data was not a possibility. Certainly some entities will achieve success. How many and who they are will determine the course of future events. The bottom line is still that no one can predict with any degree of accuracy who will fail and who will not (except that I believe governments above the county level are already largely doomed).

"9. Individual firms will fail. If we allow capitalism to work, they always will. Failure, within reason, is a good thing. In fact, we probably do not let enough firms perish. Our government "saves" some businesses that really ought to die. Our government also has many agencies of dubious value. If they do not function, it may prove that we can function as a society quite well without them."

Again, I largely agree. Your, "within reason" qualification is the obvious sticking point. Something will emerge from all this that we will call a society. I personally doubt that it will be much like the one we currently enjoy, but again, who knows? The secrecy and deception allow only speculation.

"10. You argue that Y2K is like every other software project we have faced during the last 30 years. It started late and will finish late. You also argue that Y2K is unlike anything we have faced before because it is "systemic." If it is "unlike" anything we have faced, isn't it reasaonble to expect a unique response (like the considerable increase in your public profile).

If you were arguing that one "ultra virus" could disable massive numbers of computers, corrupt data and cause serious economic disruptions... I'd agree. In fact, I think we'll see cyber terrorism in our lifetimes. If you think we ought to prepare for such contingencies, again, I agree.

To use a baseball analogy, Y2K is a slow pitch right down the middle of the plate. Everyone in the stadium knows it... and you are right, some firms will still strike out. The pitch I worry about is the wicked split-finger fastball that no one expects."

Y2K is not only like every other software project, it is like every other human endeavor in that if you start late, you finish late. And, although it may be quite reasonable to expect a unique response, nothing I've seen yet indicates such a response. The "movers and shakers" are simply "moving and shaking" with the same old tired subterfuges and techniques. The one guarantee in all this is that the results will be unique.

It's interesting that you suggest the virus analogy. I agree also. If I wanted to write the most effective virus possible; the one that would affect the most systems across all boundaries and divisions, the only one I can think of that would be more devastating than one which would cause date arithmetic to fail would be one which caused all arithmetic to fail. One which re-defined 2+2 as 3, for example. Y2K will have exactly the same effects as if such a virus had been maliciously planted in most of the world's computer systems. My experience in the computer world (and it has been borne out by countless others in print) has been that incompetence, carelessness and inattention to detail has caused far more grief than intentional damage. The origins of the Y2K problem are presently moot, the results remain to be seen.

And regardless of the type or style of "pitch", it is prudent for us all to remember that, "The mighty Casey has struck out". . .

-- Hardliner (searcher@internet.com), April 07, 1999.

Just one little snip from April 1998...

http://www.auto.com/industry/qbug23.htm

[snip]

But the Big Three are confident they will enter the new millennium smoothly. GM, Ford and Chrysler plan to finish debugging by the end of 1998.

[snip]

-- Kevin (mixesmusic@worldnet.att.net), April 07, 1999.

MR. Decker:

Automated debugging tools, aka Automated Y2000 Scanners, are able to identify a large percentage of date related coding. NOT all of the date related coding shows up, and these mostly exist for COBOL applications. There may be scanners for PL-1, or C or C++, though I haven't seen them.

The items that scanners and even retrained whiz PC programmers will miss are the programs or routines that do NO date processing, or accessing of any kind, but use a remediated data record for something else. These routines or programs wil have been scanned (by eye or computer) and will have been sealed as Y2KOK and shelved. They will be recompiled at the time of cut-over to remediated software, and, in the case of simply routines, may generate at least a W level error, which simply says that there is a small problem, and they will go on and be installed. In the case of whole programs having been sealed, they will be recompiled, generate no errors and when first run may either bring the system to its knees or generate VERY invalid data, because the program does work on data located AFTER (or to the right of) the date field. The remediators will have expanded the date field by 2 digits, and the sealed program, instead of picking up the data it expects, it will get its data minus the last 2 characters or digits, and plus the last two in the preceeding field, simply because of teh way COBOL handles data interior to a program.

This is why EVERY PROGRAM in an instalation has to be looked at BY A COMPETENT HUMAN, well versed in the Chinoiserie of COBOL programming.

Chuck, a night driver who has done some bug hunting in undocumented COBOL

-- Chuck, a night driver (reinzoo@en.com), April 07, 1999.

close tag

-- Sysman (y2kboard@yahoo.com), April 07, 1999.
.


-- . (.@...), April 07, 1999.
whoooooa.......

must be nice to be able to fix your own posting glitches! [GRIN]

"I got the POWER!"

-- lurky lurker (lurking@every.turn), April 07, 1999.

there's that smell again..

-- a (a@a.a), April 07, 1999.
A, yeah that's the smell of your breath.

Ed, I'm sorry but why does it take you so long to make a point? You are all over the field yet you didn't cover any of the bases.

-- Someone (who@knows.more), April 07, 1999.

Mr. Yourdon,

I do not have time to write a full comment on your remarks, but I did find them interesting, thoughtful and civil. It was a refreshing change from the usual sawed-off shotgun riposte other posters seem to relish. I will read your post carefully and craft a response. While we may disagree (respectfully), I hope you find the exchange worth your time.

Regards,

Mr. Decker

-- Mr. Decker (kcdecker@worldnet.att.net), April 07, 1999.

I don't know how bad Y2K will or will not be, but I'd suggest that serious economic vulnerabilities already exist. Some of these points I've previously made in scattered posts, but perhaps they bear repeating and consolidation:

1. The main stock market valuation model used by Ed Yardeni suggests that the S&P 500 index is already 29% overvalued, and that valuation presupposes an increase in corporate earnings of 15% this year and again next year, which seems wildly optimistic. In fact, average U.S. corporate profits were down 2.2% last year (down an average of 1.8% for the Fortune 500), and although there has been some improvement in earnings lately, this doesn't look to be exactly a banner year for corporate profits, either. This hasn't prevented stock speculation, of course: Amazon.com, for instance, has yet to show an actual profit, yet has a greater market capitalization than Sears has. Yardeni has also pointed out that the recent stock market advance has been very narrow; most small cap and midcap stocks have not participated. In fact, the Russell 2000 index of small caps was down for the past year. This suggests a fragile market.

2. In the past few months, Americans have been been actually spending more than their income--a phenomenon not witnessed since the Great Depression. They are doing this because, with some 60 million Americans now in the stock market, they feel "rich" because of the paper profits accumulated from stock gains. People are quitting their jobs to become day traders; in the euphoria of speculation, "easy money" seems everywhere. As an amusing experiment, last October I played the Yahoo/Web Street fantasy stock investment game (everything real except the money, alas); I'm not a Wall Street pro, but in one month I turned $100,000 into $149,500 (an ROI of 49.5%) by day trading. (Warning: don't try this at home, at least not with real money, unless you really know what you are doing and don't mind big risks and possible big losses.) Of course, that was a good month for the market generally, with the DJIA rising 10%. (Two Fed interest rate cuts helped.) My point is, millions of Americans now think it's in the natural scheme of things that they get rich by playing the stock market. Few think that any "correction" will be severe or long-lasting. History suggests no such invulnerability. It took the DJIA 25 years to recover from the 1929 crash--or 22 years from the low in June of 1932, when the DJIA was down 79% from its Oct. 1929 high. Investors caught in the crash had to wait until 1954 to see the same stock prices as in Oct. 1929.

The recent stock market euphoria has led Americans to feel very good about themselves and their future, and to spend, spend, spend, to the benefit of blue chip retailers--so, in effect, the stock market is leading, not reflecting, the economy, a phenomenon which last happened in the late 1920s. George Soros pointed out the dangers of this situation some months ago; today, the World Bank specifically called this U.S. situation "unhealthy." (Incidentally, the World Bank revised its global economic forecast down to only 1.7% growth or so in world GDP; clearly the World Bank does not expect a huge economic hit from Y2K, but then the World Bank never saw the Asian financial crisis coming, either.) We all recall when, on Dec. 5, 1996, Greenspan commented on the stock market's "irrational exuberance"; the DJIA then stood around 6300. Today, just 28 months later, the DJIA is almost 4,000 points higher, and Mr. Greenspan has very little to say--at least in public. You may draw your own conclusions. The Princeton Economic Institute has already drawn its conclusions; one of the most respected economic forecasting units in the world, PEI predicts the Dow will fall to the 5,000-6,000 level within a year or so, and might fall to as low as 3700 by 2002. This forecast is based, in part, on extremely complex analysis of global economic factors, with 10,000 or more bits of economic data fed daily into the most sophisticated economic modeling/forecasting computer in the business. (See www.pei-intl.com) PEI's "track record" in recent years has been phenomenal; it called the 1987 crash to the day. Incidentally, as nearly as I can tell, PEI has NOT factored in Y2K to any large extent; Y2K is simply not susceptible to such careful quantification. In other words, if Y2K hits hard, it may be adding to an already dire economic situation.

3. If Mr. Decker thinks that the global economic situation hasn't impacted us seriously (35% of the world is in a recession or worse), then he doesn't get out much. For starters, we have a record trade deficit. Mr. Decker needs to check out our agricultural, mining, and manufacturing sectors, in particular (though I'll grant that manufacturing has shown some limited improvement lately); many U.S. farms are already in a depression. Like Mr. Yourdon, I live in New Mexico; here it's boom times for the Albuquerque metro area, with consumers spending and spending, and retailers selling and selling; but in most of the rest of the state, down on the farm, the ranch, the oil rig, etc., it's bust times. (Again, one sees disturbing parallels to the late 1920s: euphoric Wall Street and a rural crisis.) Some basic commodities are at their lowest prices since the Great Depression; indeed, there appears to be a global deflationary spiral in place, though it's not quite as pronounced now as it was a few months ago. (OPEC has acted to protect oil prices, for example.) The ways that this unfolding global economic crisis, plus possibly severe global Y2K effects, may impact our supply lines and JIT inventory systems certainly has Yardeni and the Senate worried. (Today the World Bank acknowledged that the Asian financial crisis simply isn't going away any time soon.) At a conference recently, attended by Yardeni, Bill Gates, etc., the CEO of Sun Microsystems announced that his company may not be able to get needed components from Asia next year.

4. The world financial structure certainly looks vulnerable to Y2K. The Japanese govt.'s watchdog agency, the Financial Supervisory Agency, reported that as of last December, almost half of Japan's 19 largest banks had completed only 25% of "needed repairs" for Y2K. Given the complexity of Japanese software systems, and the huge amount of "spaghetti code" (heavily layered, customized stuff) involved, one must seriously doubt whether these banks will be ready in time. They don't seem to have allocated nearly enough money, manpower, or resources to tackle the problem adequately, judging by U.S. standards. The Japanese banking system as a whole already has an estimated 240 trillion yen ($2 trillion U.S.) in bad loans hanging over its collective head. A number of Japan's largest banks are, in fact, technically insolvent; they continue to exist only by the good graces of the Japanese govt., which has promised to prop them up come hell or tsunami. Meanwhile, Peter Gauthier (PNG) at his website has reported on the pension crisis looming over there next year (to the tune of another 120 trillion yen), which might rattle the Japanese govt. Put it all together, throw in plenty of nervous Japanese housewives whose bank accounts now are earning only .5% interest, and see what you get. And don't forget that we still have a $5.8 trillion national debt over here in the U.S., and that Asian investors are servicing much of that debt--unless a liquidity crisis back in Asia compels them to start cashing in their U.S. treasuries wholesale, that is. Then it's Katie bar the door.

There are also ominous rumbles in Europe and the UK. Last June, GartnerGroup reported that German banks (the biggest banking system in Europe) were a full year behind U.S. banks in their Y2K work. Perhaps they have caught up some since then, and I understand that the largest German banks are in better Y2K shape than are their smaller brethren, but it still appears that the German banking system might be in for a rough Y2K ride. (German banks, of course, have also been hit by Russian defaults; Germany is the largest lender to Russia.) In the UK, the Financial Supervisory Authority recently announced that twelve of London's major financial institutions may be in some Y2K trouble. The head of a Swiss software company that supplies banking software to various British banks says that he thinks quite a few British banks have been exaggerating their Y2K progress. Maybe so; maybe not--but as of now, Y2K does not look like a cake walk for the Brits. Robin Guenier, formerly the British govt.'s top Y2K man, noted that the banks are supposed to be leading everyone else in Britain on Y2K, and that Britain is supposed to be leading the world (well, the Brits are only 3 months or so behind the U.S.); if some large British banks really are in Y2K trouble, then, Guenier remarked, "heaven help us."

As for our banks, I've always considered that the great majority will be ready, if only because most got such an early start. The Fed currently claims that only 3% of U.S. banks are in any kind of Y2K trouble (i.e., noticeably behind schedule); that works out to roughly 320 banks nationwide. Some independent numbers haven't been as encouraging; for instance, Weiss, an independent bank rating agency based in Florida, claimed last November that 12% of banks in its survey were behind in their Y2K work; Weiss more recently claimed that 32% of banks in its survey failed to make an important deadline (for having all internal systems fixed) this past December. Maybe Weiss is right; maybe not. A bank interoperability test was scheduled for January; as far as I know, it has yet to take place. (If it has, the results have not been widely publicized.) This suggests some possible problems setting up the integrated end-to-end systems tests. Again, this doesn't mean that most U.S. banks won't be ready; they should be ready, with even some time to spare. But it does mean that the Y2K beast hasn't been completely killed yet, even in the U.S. financial sector which is supposed to be well ahead of the game.

-- Don Florence (dflorence@zianet.com), April 07, 1999.

Mr. Florence,

Yikes, a post I can finally sink my teeth into. I am on a time budget, so I will start you with an appetizer.

"Analysis by Mark Zandi Written February 11, 1999

The economy is performing astonishingly well. Real GDP grew by 6% annualized in the fourth quarter of last year and underlying growth is a robust 4%. The jobless rate at 4.3% is as low as it has been in a generation. Inflation is dormant. Consumers continue to lead the way. Retail sales, vehicle sales and home sales all remain extraordinarily strong. Businesses also continue to invest aggressively. Moreover, there are few signs the economy is slowing in early 1999. Payrolls expanded by a strong 250,000 jobs during January. This is similar to the average monthly gain experienced during all of last year. What makes the economy so astounding is that it is able to perform as well as it has despite enduring a seemingly endless series of unprecedented shocks. Last year it was the Asian economic crisis. So far this year it has been the crisis in Brazil and the president's impeachment trial. While the economy appears unstoppable as 1999 begins, the overwhelming consensus among economists, policymakers, investors and at Dismal is that it will weaken as the year progresses. The economy is finally expected to succumb to the shocks that have and will occur. Dismal expects real GDP growth this year of 2.5%.

The view that growth will slow this year is no different from the expectations this time last year regarding 1998 growth. The consensus as measured by Consensus Economics was that the economy would slow to 2.6% growth in 1998 from its near 4% gain in 1997. Dismal was projecting 3% growth for the year. The consensus and Dismal were clearly wrong. Why were the commonly held views regarding the economy's prospects in 1998 so overly pessimistic? What is different about 1999 that will ensure that expectations of much slower growth this year are not similarly wrong?

The 1998 economy's better than anticipated performance was largely due to a misreading of the full impact of the global economic crisis. The negative impacts on the U.S. economy of the crisis, namely the deterioration in U.S. exports and thus the trade balance, were well understood. The real trade deficit widened by a whopping over $100 billion during the year, shaving almost one and one-half percentage points off U.S. growth.

The positive impacts of the crisis on the U.S. economy, however, were much greater than expected. The immediate positive result was the dramatic decline in long-term interest rates. The global flight-to- quality into the U.S. bond market pushed 30-year bond yields down almost 200 basis points to near 5%, their lowest in over a generation. That interest rates would fall as a result of the crisis was not unexpected, only the degree to which they fell. Foreigners went from owning 20% of publicly traded Treasury bonds at the end of 1994 to close to 40% at year-end 1998.

The plunge in interest rates pushed fixed mortgage rates below 7% and set off a record year in the housing market. A record 4.78 million existing and 900,000 new homes were sold during the year. Homebuilding also soared, with more single-family homes built last year since the baby boomers were first forming households 20 years ago.

Sub-7% fixed mortgage rates also induced an unprecedented mortgage refinancing boom. A record $700 billion in mortgages were refinanced last year, accounting for just under one-fifth of total mortgage debt outstanding. The refinancing boom in turn stimulated consumer spending as debt payments fell and many households raised cash by increasing their mortgage balances. An increase in refinancing activity in 1998 was anticipated, but not the tidal wave that occurred. It was not well understood that the majority of mortgages outstanding had coupons of around 7.5% to 8% that made sense to refi when rates fell below 7% given the dramatic decline in the transaction costs involved in a refi.

Consumers also received unexpectedly large boosts to their real incomes last year from plunging energy, agriculture and import prices all resulting from the global crisis. Based on the price for West Texas Intermediate, oil prices were cut nearly in one-half last year. At their current $12 Bbl, real oil prices are as low as they have been since prior to the oil price shocks of the 1970's. The approximately 15% appreciation in the broad-trade weighted dollar also sent import prices tumbling. Import prices excluding oil prices, fell a whopping 5% last year.

The disinflation last year added at least a percentage point to real household incomes. The boost in incomes acted much like a tax cut, prompting spending increases by a like amount. Real spending rose a startling near 5% last year, the strongest growth since 1984. The unexpected surge in consumer spending and homebuilding due to overseas events thus largely offset the impact of a widening trade deficit. The global economic crisis was hardly an impediment at all to the U.S. economy last year, not the substantial drag on growth that was expected.

The economy's stellar performance in 1998 was also underestimated due to the largely unanticipated strength of the stock market. Stock prices as measured by both the S&P 500 and the Wilshire 5000, an index that includes all publicly-traded stocks, rose well over 20% again last year. Most analysts, RFA included, had been expecting the market to post returns close to their long-run average in the high single digits.

The booming stock market supported last year's surge in consumer spending via the so-called wealth effect. The wealth effect postulates that consumer spending is in part determined by household's net worth. RFA estimates that for every one dollar gain in net worth, consumers increase their spending by four cents over the subsequent two years. The wealth effect works through its impact on consumer confidence and household cashflow via realized capital gains.

The stock market's extraordinary gains have also supported soaring business investment. The cost of equity capital is at record lows, particularly for fast growing high-tech and information businesses. Government's exceptional fiscal health is also directly due to the tax revenues generated on surging realized capital gains. The stock market's outsized gains have also indirectly supported tax revenue gains by lifting household incomes, retail sales, property values and even corporate profits.

The stock market's stellar year is in large part attributable to the Federal Reserve. The Fed's aggressive actions to ease monetary policy in the wake of Russia's economic problems and the resulting near global financial meltdown, staved off what at its worst was an over 20% correction in stock prices. The Fed had little choice but to act in its role as lender of last resort. When the hedge fund Long Term Capital Management imploded the risk of a systemic financial crisis was high and rising. The resulting injection of monetary liquidity, however, quickly found its way into the stock market ensuring another stellar year for stocks and a soaring economy as 1999 began.

Given that this year has started out so strongly, it appears overly pessimistic to believe that growth will be cut nearly in half. Underlying this view is that not only will the global economy remain weak, further undermining the U.S. trade balance, but also any benefits from the crisis to the U.S. economy have been largely accrued.

After a stellar 1998 it will be very difficult for the economy to measure up in 1999. Growth is expected to slow substantially this year, posting a gain that is half that enjoyed last year. The economy will expand nonetheless and very likely extend the current expansion into the next century. This expansion will then become the longest in the nation's economic history.

Risks to this benign economic outlook abound, however. How events unfold in the global economy and financial markets in the coming months is the critical risk. A weaker global economy than anticipated will certainly put a pall on U.S. growth this year. A stronger than anticipated global economy may result in an overheating U.S. economy and a much weaker economy next year.

The overvalued U.S. stock market magnifies the risks posed by the uncertain global economy. A global economy that does not perform much as scripted could easily undermine the precariously perched stock market. A significant and sustained decline in stock prices may ultimately be too much for this expansion to bear."

Regards,

-- Mr. Decker (kcdecker@worldnet.att.net), April 07, 1999.

Simon: "If you couldn't print letters, granted you would write them out." That's funny.... have we forgotten all about typewriters? Crank 'em out on taht ol' IBM Selectric!

-- Amused (smiling@hee.hee), April 07, 1999.
Ed, I would like to comment on your introductory paragraphs. I mentioned earlier about Requirements Management and Project Management and their impact on projects. I had hoped you would have spent more time discussing your premise that projects start late and the source for your conclusions. Although you pointed out similarities between Y2K and development ("traditional"), I maintain that Y2K is still not a traditional project.

You wrote, "it is a valid point that people like Capers Jones and the Gartner Group have not made similar comments; it's more common in the software field to blame failures on unclear, incorrect, or ambiguous requirements" Let me reference the Arlie Council (spelling?) where you yourself identified 16 best practices which came about from poor management especially in the area of requirements. So this "common blame" has also been made by you. Further projects start when there is a need for a project to start. How one can infer that a project starts late can not be determined.

I haven't seen that "code fixing" for Y2K parallels traditional code fixing. Fixing implies maintenance not new software development. I haven't found that for Y2K it is 15% of the project, it turns out to be more. Further I also haven't found that tesing for Y2K parellels traditional testing. It is much less. This goes back to the requirements; for Y2K they are much less complicated. Testing will need to verify the Y2K fixes and also the functionality (but that may be more of regression testing since functionality doesn't change in Y2K).

I will leave the other points to Mr. Decker.

-- Maria (anon@ymous.com), April 08, 1999.

oops, I just went back to read that. I didn't type what I meant to type.

Projects start over time, after business strategies are defined, statement of need is made, funding is set aside, key management agrees to functional objectives, and clear technical goals (at a high level) are defined. However this is done and however long it takes doesn't matter. It's done when it's done and this milestone defines the project start. To say that the project starts late doesn't make sense. It begins only after there is agreement and other factors are in place. If something delays the start of the project so be it. That's not to say it started late, for the priorities may have changed or technologies may have changed or the business may have changed. Project start has never been a metric (I've been researching metrics for some time now).

-- Maria (anon@ymous.com), April 08, 1999.

Chuck, a night driver:

You know what you are talking about, I agree 100%, you NEVER know where the problems are going to show up!!!

-- Carlie Scott (carlie_scott@yahoo.com), April 08, 1999.

Don:

Excellent synopsis of the risk management opportunities available across all compass points. Another interconnection link has been added from the Japanese front:

The infusion of "Japanese" public money to prop up the Japanese-banking asset ratio has come from...mostly U.S., followed by European investment companies. The Bank of Japan comes in third place. The massive bond issue to repay the investment companies is still down the road this year. All told, I believe the foreign investment was close to $60 billion last week. I'll confirm the amount tomorrow.

The three largest economies, U.S., Euro group and Japan, each have different circumstance requiring different "handling." The key to navigating the shoals ahead will be, in my opinion, the NYSE. The catalyst to collective national fortunes. The structural weakness is in the manufacturing industries.

If you consider internet stocks to be "incremental profits," the base from which the increment begins, manufacturing, must remain solid. Manufacturers P & L's and forecasts are collectively weak.

-- PNG (png@gol.com), April 08, 1999.

Moderation questions? read the FAQ