Because Y2K has always been complicated, determining when we can let down our guard during 2000 is nearly impossible. But that has never stopped us on this forum before, has it?

This is my gut estimates on when we will have meaningful data from the key Y2K sectors as well as how/when we can analyze that data for a meaningful read on the ultimate Y2K impacts.

Corrections and extensions to various sectors are greatly welcomed by moi, especially by those who are experts within those sectors.

Utilities: Jan 10 to determine whether first wave of embedded failures were significant. If significant, February 15 to begin analyzing recovery curve. If insignificant, March 1 to determine significance of other impacts (fuel, internal software remediation status).

Banks: Jan 15 for first feedback, based on consumer behavior, positive or negative. Feb 15 for first read on whether database corruption is occurring. If occurring, April 1 to assess long-term damage and time to recovery.

Markets: Unpredicatable.

Telecommunications: Jan. 10 for first analysis of embedded exposure. March 1 for evaluation of time to recovery if embeddeds bite. If embeddeds are quiet, status of telecom should be clear by February 1.

Oil and Natural Gas: Jan. 10 for first analysis of embedded exposure. If embeddeds bite, March 1 for first analysis of status through June of 2000 (impossible to predict beyond June for that scenario). If embeddeds are quiet, dependencies on shipping and other political situations make April 15 the first reasonable final-status date.

Government (IRS): March 1 to determine initial damage, but anecdotal; May 1 to evaluate long-term effects. If doomed, will learn from emergency Congressional proposals in March to revamp system ASAP.

Government (USPS): February 1 to determine extent of initial delays. April 1 to evaluate improvement or continued degradation of service.

Government (Medical/SSA): February 15 for first evaluation of error rate in payments. May 1 to evaluate long-term social impacts on public health and confidence.

Retail Supply Chain: Unpredictable and varying, depending on impacts to other sectors.

Agriculture: March 1 for visible feedback on long-term problems. May 1 to evaluate long-term prospects.

Manufacturing and Trade: March 1 for visible feedback on long-term problems. May 1 to evaluate long-term prospects.

Shipping: Jan 10 to determine whether first wave of embedded failures were significant. Jan 31 to predict time to recovery.

Defense: Difficult to analyze since public information on status will be guarded.

Terrorism/war: continuous exposure through April but diminishing from a peak at rollover.

Overall (embeddeds): January 15 for first read on status of embeddeds across all sectors. If embeddeds are minor problem, Y2K will not be > 7. If they are a major problem, we will not be able to propose final Y2K recovery dates until May or later when more data is available. Fasten your seatbelt.

Overall (software): After first reports come in on sectors above, they will need to be factored together to adjust for interdependencies. Add up to one month to all dates above to account for the smoke of war (PR spin, difficulty in analysis and prediction based on analysis).

Net: Analysis of likely embedded impact should be reasonable by the end of January. A good read on Y2Ks software impact should be available somewhere between March 15 and May 15.

Note  even if Y2K proves in retrospect to be a 3, we wont be able to determine that before March 15 to April 15, since confidence in Y2K PR is rightfully low. If Y2K is heading for > 7, the PR will still insist through March that we are looking at a 3, if it can (ie, if embeddeds are quiet).

Scenarios: If embeddeds go down hard and reports from sectors on software are bad, we might know by end of January that we are looking at > 7.

If embeddeds are quiet and all else is quiet, confidence that Y2K is < 7 will (and should) grow steadily from January 31 through April 1, but take into account the Note above.

Most likely, unexpected Y2K impact consequences and PR spin will make it very difficult to assess the long-term trajectory before April 15. We should have a fairly good idea by April 15 whether Y2K is already over, will be over in several months or will be impacting us throughout 2000 and/or beyond.

-- BigDog (BigDog@duffer.com), December 14, 1999

Answers

If lights, water, phones are still on everywhere in the world on the 1st, it seems clearly not the end of the world.

If banks and exchanges everywhere open on the 3rd or 4th, we're not likely to have any financial panic.

Once the exchanges are open, count on the usual financial analysts everywhere to be checking companies for failures. If the commodities markets and stock prices don't react, I expect there's nothing to worry about.

If you don't see any shortages anywhere by the end of January, it really was a bump in the road.

Needless to say, I'll be *very* surprised!

-- You Know... (notme@nothere.junk), December 14, 1999.

Big Dog---My assessment of the situation has been considerably gloomier than yours but I couldn't agree more with your 'when we'll know' dates. Dead on and well presented analysis.

-- Get Real (gaf@mindspring.com), December 14, 1999.

BigDog: Based on how slow the problem has taken to rear its ugly head, all of your predictions may be right on. That is, you have given slack time for our bell to get rung with the embedded chips in early January, then cold cocked with systemic breakdown through first quarter. THEN, we get to assess the protracted recovery period. Excellent analysis.

One factor that could leave us FUBAR'd from minute one, the human chaos factor. If the embedded start taking sh-- down from the stroke of midnight at the rollover, all bets are off, and we go >7 and things get a whole lot worse before we get to pray they start to get better.

JM

-- Zen Angel (EndofDays@Now.com), December 14, 1999.

More industries have made decisions to stop or slow at the rollover. With the pipelines, chemical plants, trains etc. deciding to bring production up slowly this may result in a number of scenarios.

There are likely to be less initial "in your face" problems and as these industries come back online there may be less knowledge of the difficulties they are facing.

The natural tendancy is to hide failures in the hope they can be fixed before discovery. This will work in some cases. In others it will simply delay the inevitable. With many trying to hide their status, the issues could be greatly exacerbated.

Let us hope that silence is good news, not bad news.

-- ghost (fading into the@background.com), December 14, 1999.

BigDog what about Refining, Chemical Processing & Pipelines?

Refining, Chemical Processing & Pipelines - Due to be shut down for roll-over followed by restarts after Jan 1 should also quickly give clues about their impacts on other industries that are dependent upon them.

The product dependence is an immediate reaction for some chemicals.

-- snooze button (alarmclock_2000@yahoo.com), December 14, 1999.

snooze -- I guess I kind of put refining with oil. Chemical SHOULD definitely have its own category, slipped there. It's another key piece of the embedded picture.

One thing I did not make clear is that embedded impacts may vary widely be sector, depending on how embedded systems were engineered, when they were widely fielded and which, specific chips and/or updates have been made both before Y2K remediation began ... and after.

I am too inexpert to speak to that in detail.

-- BigDog (BigDog@duffer.com), December 14, 1999.

Thanks, Big Dog, for an effort that clearly took a fair amount of thought. It far outshines my paltry effort in yesterday's thread "How soon will we know how bad it is?".

In that thread I expressed my opinion that, if Y2K is going to bring us a catastrophe on the order of a 10 (for me a 10 denotes an Infomagic-style population die-off and consequent social devolution), we should have plenty of evidence pointing in that direction by January 10. In my mind, if there are widespread significant ongoing failures in the industrialized world on January 10 I expect a >6 is already clinched and a >7 is very likely.

For me, knowing if it is a >7 or not is the one piece of information I want to know at the earliest possible moment. That's the dividing line in my mind where Y2K becomes a life-threatening event rather than a life-altering event.

On the other end of the continuum, if January 10 comes and the news indicates (and personal experience doesn't refute) that the initial spike of problems has abated, banks are open and apparently in good shape, transport and communications are working with no or minimal disruption, and most manufacturing industries (petroleum, chemicals, food processing, etc.) are operating at least 80% efficiency, at that point I am willing to say a >7 is highly unlikely and economic effects will probably be the worst of it -- at least for the first year or two.

If January 10 shows something that lands somewhere in the grey area between widespread infrastructure failure and minimal disruption, we'll have to await further developments before we can rule the >7 range either out or in. I'll be as anxious as anyone to read those tea leaves.

-- Brian McLaughlin (brianm@ims.com), December 14, 1999.

Brian -- your thread was the inspiration and I agree with you 100% that "7" (plus or minus) is the key number, vague as such things are.

-- BigDog (BigDog@duffer.com), December 14, 1999.

I have been thinking "7" for almost a year now. Looks like a small consensus building here.

Great analytical job BigDog. Thanks a lot!

-- Irving (irvingf@myremarq.com), December 14, 1999.

Irving -- thx. Hoping for a 3 actually and still expecting 8.5. But, it is "past 7" where things turn nasty and, yes, I would think there could well be a consensus about that. Of course, this thread really isn't a prediction thread about the EXPECTED impacts but about timetables for reasonable discerning of impacts ....

-- BigDog (BigDog@duffer.com), December 14, 1999.

Let me respectfully suggest that you guys are ignoring the current state of the stock market.

It's possible that Y2K computer problems will themselves be a bump in the road, but will be the item which triggers the next stock market collapse and depression, which could themselves create a 7 situation.

By no means am I a stock market expert. I note, however, that it is commonly said that many companies whose stock prices have enjoyed great increases are not profitable.

I can tell you from my own observation that many more today than in the past low-educated unsophisticated types are purchasing stock market and real estate investment books. I take this to be a bad sign for a market experiencing record highs.

-- GA Russell (garussell@usa.net), December 14, 1999.

As for the stock market bubble, most of the denizens of TB2000 are well aware of it. Actually, in the rather hazy 1 to 10 scale, a 7 is usually identified as mainly consisting of an economic depression, lasting a good 3-4 years, if not a decade. Levels above a 7 usually include some measure of social violence like rioting, leading up to violent political upheavals and/or martial law.

-- Brian McLaughlin (brianm@ims.com), December 14, 1999.

Using curcular logic I'd say 3.14.

-- Squid (ItsDark@down.here), December 14, 1999.

Hey Big Dog, great post! Here is rule of thumb I will be using.

10. Eliminate the possibility of a 10 in the first five days if the utilities and phones are not in big widespread trouble. In this scenario embedded chips will be the major cause of widespread immediate shutdowns.

9. Eliminate the possibility of a 9 if, after one week, the embedded chips are creating havoc less than 30% of the utilities.

8. Eliminate the possibility of an 8 if, after two weeks, the embedded chips are creating havoc less than 10% of the utilities.

7. Eliminate the possibility of an 7 if, after three weeks, the banking system and stock/bond/market is operating at the 90% level.

6. ....and so on.

In other words, with each passing week we will have enough info to eliminate the next level down. This is the system I will use. Maybe you will not use one week intervals but ten day intervals. Perhaps someone else will use 5 day intervals. The time interval depends on you data, your experience, and your ability to predict. Some of us have many connections, tons of data, and years of business and engineering experience. Some are still not sure what and embedded chip is.

-- Tomcat (tomcat@mycall.com), December 14, 1999.