Do you suppose the following was a Freudian slip?

From the Seattle PI this morning...

Quoting Alfred Berkeley, head of NASDAQ at a speech given yesterday:

"Decimal pricing of stocks may arrive in 2000, but not until the Year 2000 computer problem is solved. Still to be worked out is whether to go to decimal pricing in penny or nickel increments."

-- Valkyrie (anon@please.net), May 19, 1999

Answers

Having just tested all the core systems on Wall Street for Y2K compliance, the boys on the street are sure as hell not going to try to implement any other significant system changes prior to next February. That's just common sense.

-- Polly (skippy@innermongolia.com), May 19, 1999.

Yeah, right, like the common sense of all those 1999 bank merger deals...

-- Brooks (brooksbie@hotmail.com), May 19, 1999.

It was the "but not until the Year 2000 computer problem is solved" that caught my eye. Seems to me he is admitting that there is still a computer problem with Year 2000. Only logical that they would put off implementing a new system until the present one is proven. This statement would indicate that they do not know if it is Ok and won't until the actual roll over which is not what they said when last I read anything on the testing they did/are doing.

-- Valkyrie (anon@please.net), May 19, 1999.

Yep, sounds interesting to me, too.

But, oddly, what caught THIS programmer's eye was his differentiation between penny and nickel.

There is absolutely NO difference to a programmer between these two.

What I don't understand is why it would be even under discussion. Why would you bother with a nickel?

When time is used up talking about trifles like these, less time is available for *real* work.

Jolly

-- Jollyprez (jolly@prez.com), May 19, 1999.

Hmmm.

Yep, sounds interesting to me, too. But, oddly, what caught THIS programmer's eye was his differentiation between penny and nickel. There is absolutely NO difference to a programmer between these two.

There most certainly is. Suppose you have a field $$$$.$$ (currency, two decimals) into which a trader puts the price he's willing to buy or sell a stock at. If it's traded in cents, that's all there is to it. If it's traded in nickels, you have to fault anything that's not a multiple of 5c.

What I don't understand is why it would be even under discussion. Why would you bother with a nickel?

The difference between (say) $100.00 and $100.01 is quite significant if you are trading stock in blocks of $10M. ($1000). And that's a small trade by some organisations' standards!

The reasons for having a "quantum" are less clear but obviously real. It's not a settlement issue, nobody would miss the fraction if one multiplied up and rounded to a single cent. However, there will be technological issues to do with performance of the systems that match bids and offers, and nontechnical issues to do with the liquidity of the market. I can't say more since I'm not an expert on big financial systems.

As for not doing this until Y2K is out of the way, that's a no- brainer that even a PHM will understand. As any sportsman will appreciate, you don't take your eye off the ball, even if you're winning!

-- Nigel Arnot (nra@maxwell.ph.kcl.ac.uk), May 19, 1999.

Perhaps, as a sometimes Wall Street computer geek, I can help on both counts.

The chairman of Nasdaq doesn't work for stockholders, but for stock brokers. A large numnber of Nasdaq brokers were convicted last year of serious securities violations involving the bid-ask spread, i.e., the difference between the sell-side and buy-side prices on stocks. This, along with commissions, is where they make their money. (Think of a bookies' vigorish.) The bottom line is, they were caught, on a very large scale, of manipulating the spread to rip off the customers. So, why the question between pennies and nickels on decimal quotes? Well, with nickels, you get a larger bid-ask spread. Any questions?

As to the Year 2000 computer problem... well, on the Street there are lots of rumors, lots of stories, etc. The recent tests have apparently gone satisfactorily, but remember that they are heavily scripted, and at very low volumes. And the trading systems are but a small part of the software infrastructure at a bank or brokerage. I just left there, and you can take it from me, they are NOT done.

There is still lots of testing underway, and a lot of the testing is not being done in a thorough way. So the powers that be might let slip that 'the problem' is not quite solved yet...

-- Brady Wiseman (bradywiseman@compuserve.com), May 19, 1999.