Top U.S. Brokerages Spent Over $1 Bln On Y2K-SEC

Updated 3:49 PM ET September 17, 1999

WASHINGTON (Reuters) - The top eight U.S. brokerages have spent more than $1 billion getting their computer systems Y2K ready, Securities and Exchange Commission Chairman Arthur Levitt said Friday.

As a whole, Wall Street has tested and retested its systems in preparation for the new year and it appears as if all systems are go, Levitt said, repeating similar remarks he made recently.

"Now there may be isolated, and I emphasize isolated, problems, but thanks to the tireless efforts of the financial industry, the disasters that some originally predicted will not get even an honorable mention in the history books," Levitt, Wall Street's chief regulator, told a conference sponsored by the President's Council on Year 2000 Conversion.

"The securities industry and its regulators have taken extensive and effective efforts to ensure a smooth transition into the year 2000."

And in case there are any problems, the SEC and other market participants are developing "information coordination centers" to monitor market activity, he said.

Self-regulatory organizations like the National Association of Securities Dealers, parent of the Nasdaq stock market and the SEC will collect information about the status of the exchanges, the broker-dealers, large mutual funds and other market participants from Dec. 29-Jan. 7, the chairman said.

"We will use this information to identify and assess problems that may arise." he told the gathering.

"Then, in coordination with other regulators and market participants, we'll be in a position to take action to solve a problem and minimize its impact on the markets and investors."

===================================== End

Ray

-- Ray (ray@totacc.com), September 17, 1999

Answers

This sounds almost impressive...until I remember how many expensive systems I've seen scrapped, because they simply didn't work. If the fixes are correct, their cost is of little import. If they don't work, it doesn't matter how much they cost...

-- Mad Monk (madmonk@hawaiian.net), September 17, 1999.

MadMonk -- You've nailed it exactly. The money is insignficant, one way or the other. Lots of bucks were unloaded on the Denver Airport and that was "going to work" smashingly (everyone knew because it had been "tested") up until the time it .... didn't.

-- BigDog (BigDog@duffer.com), September 17, 1999.

My take is that these brokerages just spent whatever it took until they could successfully pass all their tests. That it added up to $1 billion shows that they had plenty of bugs, plenty of projects, and sufficient willingness to keep spending until fixed. They also seem to have detailed contingency plans in place. I confess I was afraid the reaction to this post would be along the lines of:

Rule 1: Y2K cannot be fixed

Rule 2: When y2k is fixed, see rule 1.

Cory Hamasaki is correct in saying that IF the scope of the problem was understood, the fix was attainable. Perhaps if we collect enough reports like this one, we can identify Cory's anonymous laggard by process of elimination?

-- Flint (flintc@mindspring.com), September 17, 1999.

Yeah, it's not how much you spend, it's what you do with it. I know of a major petroleum company one of the top five in the world who took the added step of testing thier tester's once remediation was complete. What they did was they told the testers that they had another application that had slipped through the cracks. Then they inserted bad code into the app and had all four of their remediation vendors comb the app and remediate it. 2 missed the signifigant and easily identifiable problems. This forced Major Petroleum Co. to then go back and have the astute companies comb through all the apps and such that "Stupid Co. and Missedit Co." had been contracted for. This information comes straight from their Y2K program director to my ears. Now, how many other petro companies do you think had the smarts to do that? Don't bother, let me tell you. NONE that I'm aware of. Throwing money at a problem always makes you poorer and sometimes, only sometimes, wiser.

Ooops....

-- Gordon (g_gecko_69@hotmail.com), September 17, 1999.

Gordon:

According to Arnold Trembley, Mastercard did something similar for their IV&V. They deliberately introduced several y2k bugs into their sample code, and selected the IV&V vendor that found the most of them. I recall that one potential vendor found all of them. But in real life, there are always bugs. You try to minimize them.

-- Flint (flintc@mindspring.com), September 17, 1999.

Yeah Flint, I hear ya.

I wish more companies were doing IV&V. Unfortunately I know firsthand that they are not (at least in the petro industry). And I don't consider having a Big Five Accountant sign off on your remediation to be IV&V. Those guys suck.

-- Gordon (g_gecko_69@hotmail.com), September 17, 1999.

Gordon:

As you know, there hasn't been an IV&V industry, we've never needed one. So most of this independent "validation" is either the kind of sign-off you're seeing, or else it consists of running COBOL code through date-spotting utilities (themselves fairly new and buggy). Often, the firm being validated used those same utilities during their remediation, so the extra layer of filtering doesn't accomplish very much. And IV&V simply isn't available for firms that were a mess of different OS's, languages, and standalone systems with home-brewed interfaces. Like (sadly) IRS and DoD, most conspicuously.

As a result, most companies use IV&V as a PR device, rather than a doublecheck. Because in practice it isn't that helpful in most cases. Yes, there are a couple of 'real' validation houses, but they are a drop in the bucket of what's needed.

-- Flint (flintc@mindspring.com), September 17, 1999.