The real millennium bug isn't digital, it's psychological. It's not Y2K itself but everyone's perceptions of it. And in the world's capital markets - where traders sell on rumour and buy on fact - the countdown to the end of the year is already proving to be a jittery one.
At first sight this would appear strange. Overall, most financial markets professionals think computer systems will suffer hiccups rather than breakdowns over the new year. Perhaps the glitches will be worse than those accompanying the birth of the euro, but they will be far from unmanageable.
Trades may fail to settle, back-office systems may spew out corrupted data, but this shouldn't bring business shuddering to a halt. "Manual intervention is a viable alternative in almost every situation," says Jeffrey Werner, treasurer of GE Capital, one of the biggest borrowers in world capital markets.
But despite the likely modesty of the IT problem, virtually all players - investors, borrowers, and brokers - are changing their behaviour, either because the millennium is an unknown quantity, or because they expect that others will behave abnormally too. "This is all about fear," says George Magnus, chief economist at Warburg Dillon Read (WDR).
"From our point of view, the behavioural effects of Y2K are now the central issue," says Tim Shepheard-Walwyn, managing director of corporate risk control at UBS and chairman of the Global 2000 Committee, a congress of major financial institutions and authorities that meets to exchange ideas and information about the date change.
"The issue moved to centre stage around April when we started work on risk mitigation and contingency planning, because most international IT preparations are well in hand. The aim is to try to keep behaviour as normal as possible," says Shepheard-Walwyn.
But normality will be achievable only within limits. UBS's own investment-banking arm, Warburg, is typical of current risk aversion about Y2K. Its traders are banned from making transactions that settle during what have been dubbed the "red days" from December 31 until January 7.
The preceding and subsequent fortnight is designated an amber period, during which the only trades permitted to settle will be ones with a particularly big profit attached.
Other investment banks plan similar restrictions. If these major providers of liquidity are going to be so shy of counterparties for several weeks, some impact on the efficiency of financial markets is inevitable. For example, it will mean greatly reduced inventories and activities in repo markets, which will reduce the liquidity of even government bonds.
Other abnormalities have already started. The distorting effect of year 2000 fears became evident in the swap market in early August, where the absence of bids for fixed-rate interest drove 10-year dollar swap spreads to decade highs of over 110 basis points. Credit spreads in bond markets soon followed swap spreads upwards. Investors' aversion to risk and preference for liquid treasuries seemed an unnerving echo of the crisis of autumn 1998.
In addition to the well-publicized rise in swap and credit spreads, a millennium effect is also visible in repo markets, government bonds, medium-term notes and commercial paper. The phenomenon could even spread to equities this autumn. Participants in almost every market are adopting defensive policies against illiquidity, volatility, and the possibility of computer errors by counterparties.
This may be rational from an individual point of view but the collective effect is to create markets that will benefit no-one.
Or almost no-one. Two types of opportunist may enjoy the millennial markets. First, scavengers who can take a long-term view on spread movements will hunt for bargains among financial assets whose prices are suffering from temporary flights to liquid government paper.
Second, hedge funds, investment banks and aggressive fund managers may see an opportunity to take a directional view on unusually thin markets - and then push them in the desired direction. This autumn will be a test of how much appetite for such risky trades remains among macro funds and proprietary traders, following last year's overleveraging.
There are two scenarios for the final months of the millennium. In the first, capital markets become frantically busy in early autumn as last-minute positions are squared, fall asleep from November to January, and spring back into action once participants see that everyone's computer systems have coped with the date change.
During the sleepy period, few investors or traders will take many directional views. Asset prices will be volatile but dips or spikes will be artificial, merely reflecting wafer-thin trading volumes. The idea is that if you sit on your assets and wait, stability and sense will return to prices during January.
The second possible outcome of millennial fears is more worrying. Instead of sitting neutrally on their positions and awaiting the return of normality, fund managers could be forced to liquidate securities holdings cheaply by retail clients' demands for cash.
Meanwhile, investors' discrimination against parties whose Y2K-preparedness is unclear, such as emerging-market banks, could lead to big capital reallocations around the world. Opportunistic trades by the remaining risk-takers could deliberately worsen the volatility of assets that don't count as safe havens.
Another emerging-markets crisis could ensue, with phenomena such as high credit spreads spilling over into early 2000. And abnormal millennial behaviour could even trigger the long-awaited correction in US stocks, leading to pain in equities everywhere.
Analysts disagree about which of these possibilities - comatose markets or seriously turbulent markets - is more realistic. The majority expect the millennium to affect the level of securities trading and issuance more than the direction of prices. But as Peter Fertig, chief fixed-income strategist at Dresdner Kleinwort Benson, says: "We are in a bear market which has fundamental reasons, but to some extent this has to do with Y2K, and we don't have any experience with this problem."
If capital markets grind to a halt, it will affect the business flow of brokers and the investor returns but the situation could be more threatening still for borrowers. Werner of GE Capital believes that IT systems of major market participants will be ready for the date change, but the danger from Y2K will come "if investors are going to close down their books, and if investment bankers plan on running down their secondary inventory".