Wednesday, August 11, 2010

You Cannot Force People To Make Markets

At the risk of beating a dead horse, I want to come back to an essential point that is necessary to remember in any discussion of market structure reform.  This morning I praised Floyd Norris for an interesting article about how MBIA's loss expectations have evolved.  Tonight I give him a thumbs down for his analysis of "liquidity providers" versus "market makers."  Norris writes:

"I have heard people from Wall Street say that it is unreasonable to require market makers to make markets rather than simply post “stub bids” of one cent when they do not want to take part.  One even suggested that the solution was to bar retail investors from placing orders to sell at the market price, lest such orders hit a market unwilling or unable to handle them at reasonable prices.  Such an order should have a 3 percent range of allowed prices, so that the order would not be executed at an unreasonably low price, it was said.

Give me a break.
This could never have happened a decade or two ago. Then markets had market makers, who were required to maintain bid and ask prices.  There were primary markets that set prices.  Prices could plunge — see 1987 — but not like they can now without anything really happening other than a breakdown of Wall Street systems."

Since I've addressed these topics at length previously on this blog, I'll just copy&paste the comment I left on Norris's post, and refer you to my prior writing on the subject:

"Floyd - the point of preventing market orders is to protect unsophisticated retail investors who don't understand what "market" means. I'm all in favor of it, because I would guess that 99% of the time a retail investor sends a "market" order they don't really mean "market" - they assume that they will be filled near the current price, which is a faulty assumption.

As for other possible solutions: you can't force market makers to step in front of a falling market. It's impossible. It doesn't matter what you call them - market makers, specialists, liquidity providers - markets can still crash.

I'm not sure why you claimed "This could never have happened a decade or two ago" and then provided evidence that it did - 1987. Markets have always crashed, and they will always crash - it cannot be legislated away.

Note that the markets recovered almost instantly from the May 6th "flash crash's" temporary lack of bid-side liquidity - the same cannot be said for the 1987 crash which was presided over by real live market makers."



Anonymous said...

Good for you for trying to bring reason to this argument. Sadly, your comments didn't make it onto the times web site.

One other point that doesn't come up in this argument very often -- the retail sellers all fighting for the exits are much much faster than they were in the seventies and eighties. Getting a trade into the market in the 1987 crash was a challenge even for the professionals because the order entry systems were so bogged down (read the section on this topic in Daemons of Our Own Design). Retail investors were trading by phone in those days. Keeping an orderly market in a crash now is much harder.

Caveat B said...

I posted a comment yesterday that didn't take either.

Maybe he's adopted Krugman's comments policy (i.e. stop them, after drowning in a sea of counterfactuals).

"Cassandra" said...

Do you see merit in the Japanese approach of structurally preventing freefall due to market imbalances by sunshining aggregate bids and offers and NOT allowing trade until the imbalance clears? This may serve the fairness issue of trying as best as possible to allow everyone to trade at a clearing price, as well as allowing time for liquidity to be flushed out in response to information dissemination.

Kid Dynamite said...

Hi Cassandra - you tell me - do you like how it works there? You're in Japan right?

I would think that it certainly isn't ideal when you sometimes get locked out of trading a stock that is halted for days at a time.

in general, i'd certainly prefer the option of liquidity - the trading halts we're trying to implement here are kinda (not exactly) a little micro-scale of what Japan does, only on a shorter scale, and without the requirement that the imbalance clears. I think I'd rather have a series of 5 minute trading halts in succession than a day without trading.