TraderFeed, and bzbtrader have been talking about how large orders are split into smaller orders. TraderFeed wrote:

In other words, if you were to simply look at trades and trade volume, you’d conclude that small traders were dominating the marketplace. The reality is, however, that large traders continue to hold sway, but have succeeded in disguising their presence.

This is not new, and in fact my broker supports it and calls it the iceberg order. I talked to some Investment Banking folks and they told the iceberg order is pretty run of the mill type stuff.

What the iceberg order does is the following:

  1. Show a small order on the order books.
  2. When small order is matched causes a larger number of shares to be traded.
  3. The larger number of shares NEVER, and I repeat NEVER show up on the order book.

So I asked, if they never show up in the order book do they show up at all? The answer was that yes they showed up in the official exchange statistics because they have to.

What this means is that if you back test taking volume into account then your algorithm will not work trading live. In fact taking volume into account at all during live trading is pretty darn useless. You only know what the volume is AFTER the fact and since we don’t have time machines volume from a trading strategy perspective is noise!

Of course you could investigate the correlation between order book, and the actual after the fact volume. I say that this could be interesting because most of the trading is done by programs, and programs don’t change that often. Thus there will be repetition and there is an statistical arbitrage moment. I just gave you readers a free strategy.

I don’t use this strategy because it is scalping and I just don’t have the bandwidth, speed, or the desire to implement yet another strategy.