Monday, October 09, 2006

Tracking the Large Trader


In my post on the structure of market reversals, I presented a common sequence of price and volume events that occur at market transitions. This sequence--and the notion of transitional structures as patterns that present themselves at multiple time frames--was also elaborated in my recent Webinar.

In the chart above, we see a transitional structure from Friday's market. Notice the distinct elements: the momentum extreme featuring high volume and a high proportion of trade at the market bid; further selling and a price low on lower volume and more modest downside momentum; a significant bounce from the price lows and then an exhaustion of sellers, who can no longer push the market to new lows; and finally a reversal and influx of buyers.

What makes the Market Delta chart above unique, however, is that the volume data are derived solely from trades of 100 contracts and more. Indeed, the numbers in the bars are not volume at all, but rather the number of large trades that are being transacted. While not all large traders necessarily transact their business in such large increments, we can rest assured that small traders are not flinging hundred-lots around. By printing only trades from large traders and seeing whether those trades--and only those trades--are predominantly occurring at the bid price or at the offer, we can see in a rather transparent way whether large traders are aggressive to the buy or sell side.

Two pieces of data from the chart are important:

1) In the exhaustion phase, note how there was an influx of large trades hitting the market bid. When this selling could not push the market to new lows, those large traders had to cover their shorts, contributing to the subsequent runup. Finding spots where large traders are pushing the market one way and cannot succeed very often produces useful short-term countertrend trades;

2) After the runup, also notice how the number of large trades expanded, with many occurring on the bid. This influx of large sellers is common at short-term market tops and can aid order-flow oriented traders in taking profits (or at least in not jumping aboard the tail ends of moves).

Finally, to amplify a recent comment from the Weblog, since volume is so well-correlated with volatility, tracking the number of large trades in the market gives us some idea of the short-term opportunity for the day timeframe trader. We can readily see large traders enter and leave the markets and we can note the price levels at which they shift their participation.

If less than 10% of all trades control over 90% of all volume, why not focus on those and separate signal from noise? One of my current research projects is constructing market indicators derived solely from large trade data. I hope to report more in coming weeks.

14 comments:

yinTrader said...

Hi Brett

It is Columbus Day at your end and as such it is a holiday although certain sectors of the market are open.

Quote
If less than 10% of all trades control over 90% of all volume, why not focus on those and separate signal from noise? One of my current research projects is constructing market indicators derived solely from large trade data. Unquote.

Following from your research, I think I should not trade today as volume will not give a true indication today.

Brett Steenbarger, Ph.D. said...

Hi,

I like to think of volume as the *energy* behind trade setups. When large traders are mostly absent from the market (which we can see by low volume compared to normal volume for a given time of day), the setups have little energy--and hence reduced opportunity. As I'm writing, we've been open 45 min and trading in a 2 point range. I, personally, don't trade those markets. Too little opportunity relative to trading costs and risk.

Brett

Simple Trader said...

I was reading an article that talked about liquidity providers like LiquidNet and dark rooms that either distribute the large orders throughout the day using something like VWAP or perform block transactions outside the exchanges.

Also I am assuming that most large institutions and traders trade the larger contract and primararily use the eMini for arbitrage opportunities. And supposedly 50 to 70% of all volume is program trading with a large amount of arbitrage across the various markets leading to historic low volatilities across several markets.

So I wonder how accurate Volume is in this age?

Brett Steenbarger, Ph.D. said...

Yes, thank you for the excellent observations. Institutions do, indeed, distribute their large orders into pieces. There are sophisticated research efforts that identify these pieces and even reassemble them for analysis. I have not found that large traders necessarily use the big contracts for their business. The liquidity of the screen has taken much of the business from the floor. And, while program trading currently accounts for 45% of NYSE volume, it is possible to sort out program trades from non-program participation, in part by identifying institutional behavior at times when the futures trade out of line with cash, ETFs, etc. H.L. Camp, for instance, provides such a service.

So, yes, volume analysis *has* gotten more challenging. I do find, however, that it is possible to go underneath the market hood, so to speak, and gain a sense for what large traders are doing.

Brett

D TradeIdeas said...

Here is one example of how Trade-Ideas tracks the large trader. Using block trade alerts and large volume on the bid and ask, Trade-Ideas can identify in which stocks the larger traders are participating. Tools like Market Delta can then provide more detail on the specific stock.

We call this strategy: Big Block Trades on Liquidnet, POSIT, and other Networks

Brett Steenbarger, Ph.D. said...

Thanks, David; that sounds like a very useful alert system--particularly over time as users get a sense for normal vs. abnormal amounts of large trader participation in their particular stocks or ETFs.

Brett

John Cook said...

Brett,

Are you filtering for 100 emini contracts or 100 full S&P contracts?

John

Brett Steenbarger, Ph.D. said...

Hi,

All my data are from the e-minis.

Brett

Roshigary said...

I don't know if this will help, but check out Pascal Willain's work on Effective Volume in Elder's book "Entries and Exits." You can find Willain's patent proposal on the web, which provides more information. Maybe you can integrate some of his work into your research.

Brett Steenbarger, Ph.D. said...

Thanks; that's very helpful. One other correspondent mentioned that research to me, and it looks very promising. I appreciate the mention--

Brett

Nani said...

Hi Bret,

I for one strongly bleieve the corelation of large trades and price reversals.

Looking forward to your research on large trades and price movements.

Regards

Brett Steenbarger, Ph.D. said...

Hi Nani,

Thanks for your comment. My research suggests that large trades are more likely to reverse price moves in local dominated markets than in markets dominated by "paper". Who the large trader is matters quite a bit.

Brett

icoa said...

I'm intrigued by your recent and past posts concerning finding visible signs of large trader activities using sources like MarketDelta. I'm also curious about HL Camp and Company Program Trading research. As a relative novice, how quickly do you imagine I might grasp and benefit these services?

Eyal said...

Hi Doc Brett,

It looks like the webinar link is dead. Do you have by any chance an updated link or maybe a recording somewhere?

Thanks
Eyal