Sorry to keep you waiting this week! When the amount I’ve got to do confronts the limited intellectual capacity I’ve got, what happens is stuff takes longer.
I just splashed through verdant Cincinnati, where spring runoff sluices from the hills and household names adorn the big buildings downtown. Americana. Had a fine session there with a great group of executives for one of those names about how trading works today.
Three times this week in data for clients we saw profound disparity between the share of trades executing on displayed markets and those matching up elsewhere, perhaps between brokers directly or in the relative solitude of dark pools.
An agency trader who’s as smart a guy as you’ll find wrote in an email to me last week: “I can tell you that the percentage of non-displayed trades keeps getting larger. We are constantly being forced to route to dark pools because our displayed liquidity is always last to be executed. Price discovery is being damaged because true supply and demand is being distorted by internalization.”
What’s that mean? Suppose you ran a chocolate Easter-egg store. Each time you posted a price on your eggs, you’d see a whole bunch of people with eggs they’d just bought from other vendors. You’d figure out soon that the way to sell eggs is to remove the prices and wait for somebody else to signal a price so you can undercut it.
An “agency trader” only works fiduciary orders for clients. My friend is saying that if they display offers to buy or sell shares, the orders end up last in line. Instead, they have to take chances on prices in the dark, and those chances are rising.
These trades might be in your stock. But if they don’t occur on your listing exchange (where about 30% of trades happen), you won’t see the data. Still paying the same fees for listing? Maybe public companies should ask that listing fees be reduced to reflect the percentage of volume at the exchange?
On April 6, a client had good news. Big brokers reacted with buy orders. But get this: 86% of orders were either in the dark or high-frequency trading (or high frequency trading in the dark). The gains were gone in two days.
Why is this happening? Reg NMS decrees that trades must match at the National Best Bid or Offer. Suppose you gathered people of all shapes and sizes and told them that they’d be judged on foot speed. Soon, the only people left would be sprinters.
Which brings us to the up/down limit rule on stock prices that the SEC and the exchanges have agreed to implement. Rather than circuit breakers, all the places where stocks trade will impose “price bands” on securities to have them pause if they move a certain percent from the average price in a short period of time.
Keep in mind, these are your shares.
Taking a timeout so cooler heads prevail makes sense. But controlling the movement of prices because they’re uncertain is not free-market thinking. We’ve tried this with our economy too, through the Fed, which intervenes to smooth out prices whenever they fluctuate outside defined bands. Ask yourself: is it working?
Finally, who benefits from forced behavior? Think about it. Exchanges make money on transactions and data. In 2010, the NYSE had gross revenues of $3.5 billion on transactions and data, and $422 million from listing fees. But transactions aren’t a margin business; they’re a volume business. Exchanges charge you about $0.30 to “take” shares, or buy them. They’ll rebate trade costs by $0.29 if you “make” liquidity.
Traders making and taking lots of volume generate a lot of transactions. The by-product of transactions? Data. It’s like splitting a water molecule and getting hydrogen and oxygen. Sell the hydrogen and oxygen, that’s a margin business.
The hydrogen and oxygen that come from trading transactions is data. Thus, much of what is occurring on exchanges – and which will occur in greater quantity if exchanges merge – is noise. Incentivized trading that produces a valuable by-product for exchanges.
But what’s it done for you lately, public companies?
Plus, what works best for machines doing things furiously at high speed are KNOWN PARAMETERS.
If there are limits on price-movements within time parameters, even I could build an algorithm for it. I know my risk limits. I can program my algorithm to almost always be right. I can program it to find related issues with varying liquidity and trade them up near the parameters to create arbitrage opportunity.
But value is even less certain than before. We have utterly evacuated the most basic economic principles this way: Scarcity and choice.
Plus, it’s not investing. It’s a sort of financial Sudoku.
