Saturday, October 27, 2007

Trade Like a Card Counter

In two recent posts, I have outlined baseline odds of prices hitting various benchmarks, including the prior day's highs and lows and the previous day's average trading price.

What makes trading interesting is that these odds shift dynamically: as markets move, so do the odds of hitting those benchmarks. It is the inability of market participants to adequately update their outlooks in the face of recent events that creates one important source of short-term trading edge.

When trading is compared to gambling, the implication is generally negative: that traders are little more than people who roll dice in hopes of a big payout. There is another side to gambling, however, typified by the card counter. The card counter, dealt a hand, is aware of the odds of winning with those particular cards. The counter also follows which cards have been dealt already, dynamically updating the odds that new, favorable cards will be forthcoming. It is this knowledge of odds--and the ability to update them in real time--that makes card counters so formidable that they are banned in many casinos.

Yesterday's market is like a set of cards dealt to us. Then we get a new card with the overnight market. The market open provides yet another card. All the while, the new cards either improve or fail to improve the odds that we'll hit target prices as the day moves forward.

As this analogy makes clear, a major source of trading edge becomes the decision to not trade. Just as a professional poker player will muck many hands when the odds are unfavorable, giving up a small amount to preserve the opportunity to bet large when circumstances are more favorable, the professional trader does not need to trade. Rather, the trader bets when the odds of winning are enhanced.

Such a trading approach emphasizes the exit--the target price--as well the entry. Yes, it's important to get as good an entry price as possible, but it's knowing the odds of hitting those benchmark prices that ultimately define the good trade idea if you're trading like a card counter. By controlling the bet size--not going "all in" on any one idea and risking ruin--and by exiting as soon as market events take the odds out of your favor, you let probabilities work in your favor.

Notice that this is a major reframing of stop-losses. A stop loss level is not defined by how much you're willing to lose. Rather, it's defined as that point at which the odds cease to be in your favor. The poker player will draw a new card that adds nothing to the hand and folds shortly thereafter: the updating of probabilities tells him to stop. But if you don't know the probabilities to begin with, it's hard to hold positions until the benchmarks are hit, and it's difficult to know when and where to stop playing.

Every day in the market offers us a few hands to play. To win the tournament of trading, we play many, many hands. Consistency--knowing and following the odds--distinguishes the professional gambler from the guy feverishly feeding a slot machine. Perhaps it's not so different in markets.


Handicapping Odds in Trading


Brandon Wilhite said...

I wonder if it would be helpful to track these kinds of statistics all the time, but to only use them for trading when they are telling us something. For example, maybe there are certain stretches of time in the markets when small gaps give us good information, and other times when the statistics about small gaps give us no useful information. Personally I've only ever looked at statistics like these over long periods of time. I've never attempted to track how the usefulness of the statistics changes over time. Any thoughts?


Brett Steenbarger, Ph.D. said...

Hi Brandon,

You raise a great point. It may well be that measuring gaps and market moves in volatility units (i.e., as a function of average daily price range over a given lookback period) would be most effective over the long run. What makes a large or small gap might depend upon the volatility of the market at the time.


High Probability Trader said...

Dr. B,
You give a lot of stats on different setups in the market (ie market is down 5 days in a row or market has had 3 inside day),,what I'm getting to is, do you have these stats organized so that traders can look at them in a list, or do you plan on continuing to use random posts of stats based on whats happening the given week. It would be nice to see all of the work you've done in an organized format.(you may have already done this, so excuse me if I'm wrong, and I am aware of the search feature on your blog). Half the time I read one of your stat posts I see no strong evidence to have a bias for the given trading day, however the posts on open interests, blog traffic, and your usual stuff on trader psychology is great, also the post on hedged trades DIA/IWM was good. I haven't seen you talk much about the effect interest rates have on the performance of big vs small cap, which is something Fisher talks about in his book and I think is interesting to look at.

Sabretache said...


That's a good summary of my own approach to day-trading about half a dozen stock index and commodity futures. IMHO the term 'stop-loss' has thoroughly negative (and hence unhelpful) psychological connotations. Rather both my trade entry and exit decisions are always framed by the same question: "As of the last close, will this instrument trade at price 'A' before it trades at price 'B'?" With the probability being calculated real-time. A & B may be varied as a trade progresses, but the form of the question remains the same. The combination of inputs capable of providing a realistic probability for the 'yes/No' answer are of course as legion and tricky as ever but, per your post, the approach is more analogous to the card-counter than the blind slot-machine feeder. Fact is that the odds can change VERY rapidly indeed and certainly more rapidly than the time required to penetrate those perennial and debilitating hope/fear emotions.

Brett Steenbarger, Ph.D. said...

Hi High Probability Trader,

Creating a compendium of tested historical patterns would be an interesting and worthwhile project, though well beyond the scope of a blog IMO.

My goal with the blog is not so much to do people's research for them, but to illustrate the kinds of patterns that may and may not have favorable expectancies. It's my hope that this can help guide traders' own thinking and research going forward.

For traders wanting daily historical pattern information relevant to the current day's trading, check out the site and the Market Tells newsletter from Rennie Yang. Both are subscription based, but provide a wealth of data daily.

And yes, thanks, the topic of interest rates and sector performance is an excellent one; great suggestion--


Brett Steenbarger, Ph.D. said...

Hi Sabretache,

That's a *great* framing of the "stop-loss" issue, and one that is much more in keeping with probabilistic reasoning. Thanks!


Ferro2 said...

Hi Dr. Brett,

I couldn't agree more. The hardest part of my system is to wait for entering a trade until I get a green light from one of my signals...

When the markets go up and I'm not engaged, the urge to enter upfront becomes really huge, and I start thinking that I *could* build up a small position just to participate in the move; the main reason, why I haven't done so until now is that such a trade would nullify the whole system that I've been working on for the past five years and I'd hate to trash that massive amount of work.

What also helps a lot is the fact that my trading decisions are made without exceptions at weekends, when the markets are closed, and so there's always enough time to calm down and revert to a rational state of mind - this also helps me when positions start going against me and the emotional pressure to abandon them prematurely becomes intense.

Many thanks for your outstanding work & best regards

Brett Steenbarger, Ph.D. said...

Excellent observations, Ferro2; thanks for the comment. I like to mentally rehearse entry and exit criteria before and during the trade to cement good execution. Indeed, it was my failure to follow through on my mental rehearsals that led to the sloppy trade I wrote about recently. So much of good trading boils down to consistency!


Martingale said...

Love this post Brett and believe it will be particularly valuable to neophytes who often lack perspective.
There is nothing like sitting at the blackjack table when the count is massively in your favor and there aren't many cards left in the shoe.On its journey to zero, where it always starts & finishes,all those dealer breaking fat cards will have to come out( I'm talking 20 yrs ago forget it now)You know that now its time to play "basic strategy" and put the big bets on. Losses will come but the winners will swamp them.Put those same bets on at a random time and you will lose approx 8% on turnover.To me this is the difference between professional traders and amateurs who have memorized every candlestick pattern but just cant cut it. Finding an edge in Blackjack is simple.Learn basic strategy & learn to count until its second nature - You're there.Trading the markets is more difficult.You wont be thrown out for winning but in the search for an edge you will be chasing a moving target.