Got this off of the ECU board. VHF posted it. He is a valued poster over there.

This is a surface level walk through the high frequency trading programs activity in the markets. This says to me that I better get better aquanted with this form of trading. I read that it accounts for over 70% of the daily volume in North American Markets. Not sure that # is accurate though. Given that I make under 50 buy and sells in a year most of my trades are well planned both on the buy and sell side but, I sure don't recommend making any trade "@ Market". I know Funky crap has been going on in the markets for a long time but I think there is no duobt now that things are set inplace by design or other wise that will increase to extream volatility.

Deno G.

A Big Red Flag

Graham Summers

June 2, 2010

Since July 2009, I’ve been harping about high frequency trading programs (HFTPs) and the risks inherent in letting the financial markets be dominated by computer trading (as if October 1987 and Long Term Capital Crisis didn’t prove this point already).

If you’re unfamiliar with HFTPs, for starters you should know that most indexes or exchanges offer a ¼ of a penny rebate for broker dealers who put in orders. So simply putting in an order, even if there’s no profit to be made, can make a ¼ of a penny in profit.

The way these trading programs work is as follows… Let’s say an institutional investor has put in an order to buy 15,000 shares of XYZ company between $10.00 and $10.07. The institution’s buy program is designed to make this order without pushing up the stock price, so it buys the shares in chunks of 100 or so (often it also advertises to the index how many shares are left in the order).

First it buys 100 shares at $10.00. That order clears, so the program buys another 200 shares at $10.01. That clears, so the program buys another 500 shares at $10.03. At this point one an HFTP will recognize an institutional investor is putting in a large staggered order.

The trading program then begins front-running the institutional investor. So it puts in an order for 100 shares at $10.04. Whoever was selling shares to the institutional investor would obviously rather sell at a higher price (even if it’s just a penny). So this investor sells his or her shares to the trading program at $10.04. The trading program turns around and sells its shares to the institutional investor for $10.04 (which was the institution’s next price anyway).

In this way, the trading program makes ½ a penny (one ¼ for buying and one ¼ for selling) AND makes the institutional trader pay a penny more on the shares. It doesn’t seem like much, but do this billions of times a day and you’ve got some serious profits.

As Goldman and JP Morgan’s latest trading records show (both had flawless quarters without a single losing day), this activity can be absurdly profitable. However, for those on the other end of these trades (everyone), this is essentially stealing. The only reason that it is tolerated is because HFTPs are supposedly supplying liquidity to the markets.

How’d that work out on May 6, 2010?

As far as I can see it, there are two core problems with HFTPs (aside from the stealing, front-running, and other illegal activity that is permitted with impunity). They are:

1. There is no requirement for them to remain active in the markets
2. Because they’re based on correlations, they inevitably will blow up

Regarding #1, HFTPs publicly claim that they provide the benefit of liquidity to the market. But the reality is that there is no law that requires them to stay “in the game.” So when the market dives and their algorithms blow up, these guys can simply pull the plug, resulting in a bid-less market: a market in which no buy orders are issued.

A market like this:

Regarding #2, ultimately HFTP algorithms (no matter how complicated) are based on correlations (the relationships between various investment classes). That is, these programs are based on linear thinking (if A, then B type thinking). Because of this, invariably they all blow up because at some point or another, the correlations they follow decouple and they end up on the wrong side of the trade.

When this happens, you begin to see wild volatility with massive up and down days. We saw this in August 2007 when the credit markets jammed up. And we’re seeing it again lately.

Look at the massive spikes and collapses in the S&P 500 futures markets in the last week.

We’re talking about 3-5% moves occurring almost on a daily basis. This is exactly the kind of market action we saw back in mid-to-late 2007 when the markets began to move wildly as various quant funds got caught on the wrong side of trades and were forced to buy or sell entire chunks of the market wholesale.

This, taken along with what happened on May 6, 2010, should be a BIG RED FLAG to all market participants. You don’t get 3-5% moves every day based on fundamentals. You get this when the market is busted, overrun by parasitic trading programs whose correlation-based algorithms are blowing up.

In light of this, the following are clear:

1. Volatility is not going away any time soon.
2. The market has very likely entered Round Two of the Financial Crisis.
3. Long-only investors should seriously consider taking their money off the table.

For those of you who are passive investors (the long-term buy and hold), it is very likely a good time to be moving to cash or high quality blue chip holdings. The insane rally of 2009-early 2010 should be seen for what it is: a gift from Ben Bernanke before the second wave of deflation takes us back to the March 2009 lows. I suggest talking to your financial advisor about getting out while the getting is good.