What is My Opponent Holding?

Having spent far more hours playing poker than sanity allows, I like to think that I have more than money to show for it. I hope that I have come away with lessons that apply to business, investment, and life in general. That may be a stretch, but I cling to that illusion.

I am constantly trying to figure out why the best players are that good. It’s easy enough to see why the bad players are so bad. They bluff too often, call too often, don’t raise to protect their hands, don’t control their emotions, and have no understanding of position. You can become a mediocre to good player simply by addressing these problems, but great players have something else. As a matter of fact, even the great players still have some of these problems on an occasional basis.

Something else makes them great. I don’t think it is that the best players have more heart, more brains, or more luck, though everything does go easier when you have more advantages on your side. I think what separates the great players is their awareness that poker is a relative game. The great players spend most of their attention focusing on what their opponent has. Bad players often don’t think at all about the opponent’s cards, good players think about them at the time of decision, and great players think about them even when they are not in the hand. Every time you show a hand down, you are giving a great player another weapon to beat you with.

The great player knows that there is no greater advantage in poker than knowing your opponent’s cards. We could all be poker millionaires if we knew what our opponents have and they don’t know what we have. One of the best examples of this was a couple of years ago in the main event at the World Series of Poker. Sammy Farhar called a massive pre-flop raise early on the first day with only a pair of deuces. He knew that the nervous amateur making the bet would not have made that bet with anything other than a pair of aces. When a deuce came on the flop, Sammy got the rest of the amateur’s chips. I’m don’t think that the implied odds that he got really justified Sammy’s call, but know just exactly where he stood was a powerful incentive that he could not resist. He made it pay off.

Daniel Negraneau is always talking about what his opponents have when he is heads-up in a hand. I am sure that he only refrains from discussing what other people have in their hands in other situations because that would be outside the etiquette of the game, but I am sure he is thinking about it all of the time. I find Daniel the most instructive of the leading pros to watch for just this reason. Listen to the questions that he asks himself.

Why did he bet that much? Why did he check? What did that card do for him? How likely is it that my opponent is bluffing? Is this a guy who only bets the nuts? How nervous is this person about this bet? How many chips does this opponent have left? What is this opponent likely to do if I raise?

Daniel doesn’t always come up with the right answers and he’s quick to notice when he is wrong, but he is always asking and answering the questions. That gives him many chances to be right than he would have if he never thought of those questions.

It’s just like that in business and investing. Asking all of the right questions doesn’t guarantee that you will win, but it does increase your chances dramatically. There are never any guarantees, but there are lots of edges if you look for them.

Just spend a few minutes thinking about the last few years at some companies disappeared.  They disappeared not because they came up with the wrong answers; many disappeared because they simply didn’t ask the right questions.


Your Linked-In Connected Blog

One of the things that you can do on your Linked-In profile is link to a blog. I started this blog on WordPress.com because Linked-In offered it and said that it would show up on my profile page.

It would be really nice if Linked-In Updates would report when I post another entry, but they don’t. You actually have to go to my profile page to see that I even have a blog, and that doesn’t show to people outside your network unless you have an upgraded (not free) account. Upgrade to the business account, and people will see your blog on your profile if you so desire.

Always remember that even if no one reads your blog for months at a time, somebody will when you go to look for a job or attempt to sign up a new client.

That means that you don’t write about politics unless you want to make about half of the people out there want to avoid you for your political leanings, you don’t want to talk about religion unless you are in the religion business, and you don’t want to talk about your partying life unless you are selling your partying life. You wouldn’t talk about your bad habits in an interview. A blog will get to your interview before you do so keep that in mind.

In fact, one of the nice things about having a blog is that it is like an interview where you are allowed to ask the questions and give the answers, but someone who did not participate in that process gets to judge it. Make sure that you ask questions that are interesting to your perspective interviewers or customers, and make sure that you provide thoughtful answers. They can’t always be the right answers, but they can always be thoughtful. And because it is in writing, you don’t have to show your thought until it is complete.

If you can write, you should. Everybody needs people who can write, and the more technical your field is, the more your ability to write will be appreciated. However, you do need to remember that your Linked-In persona is the one you are taking to work. Your Linked related blog is not where you want to rant and rave.

You should only write about things that relate to the way you earn your living or want to earn your living. That gives lots of us a lot of latitude as long as we are creative. For example, I am a process professional who builds software, directs research, and markets that research. I’ve been an entrepreneur for forty years. I can write about anything from at least one of those perspectives, but I do have to couch it in that perspective.

For example, I love to play golf, but I’m not in the golf business. That means I can talk about the marketing of Tiger Woods (a topic I find fascinating), but my Linked-In related blog is not a place to discuss potential cures for my hook. If I was in the golf business, I’d have a little more latitude. If I was a golf architect, I could do course reviews or blogs on hole design. If I was a club maker, I could write about swing dynamics. The point is that you should only write about things that have a chance to help your business.

So how does this and other articles on Linked-In fit in for me? Linked-In is a tool that I am trying to exploit to its fullest extent. Part of my role and part of my interest is always on marketing, and Linked-In is a marketing tool. You have to understand marketing to some extent no matter what your job is because unless someone is buying, you may not be getting paid for long.

A Brief History of Our Application of DoE to Automated Trading Systems

Although we have been believers in Design of Experiments for nearly twenty years and builders of automated trading systems for nearly ten, we never put the two together until recently. A designed experiment simply our trading system would have taken years as recently as three years ago. As a result, we did like everybody else does. That is, we set our system parameters using intuition and experience, and tweaked those each time we reviewed the findings from any test run. We made significant progress in that way, but it was plain to see that we could learn more faster.

We started applying Design of Experiments to automated trading systems about a year ago, but the tests were taking a long time. We were working with intraday data, and that is really dense. It took many months to complete a study that used just two years of market data. Even at that cost, we were able to run a couple of studies and became convinced that the approach was valid. Unfortunately, the systems were proprietary and secret so we were having a little trouble talking about our approach in any detail without betraying a confidence.

A couple of months ago, Ernest Chan suggested that we test something from the book Short Term Trading Strategies That Work. We needed to talk about something other than our own systems. There were two major surprises.

The first was the speed of the tests. Our tests for the options trading systems that we built took six hours for each trial for each expiry (month). An eighteen month study takes about three months running ten of these tests at a time in parallel. The strategies in the book used only end-of-day data. As a result, a test on a twenty-year set of data takes one quarter of a second! We had the study completed, and article on it accepted by a publisher in less than a week.

This is mind-boggling because in most experimentation the cost of the trials is the overwhelming majority of the cost of the experiment. The challenge of most of Design of Experiments is to get as much information out with as few trials as possible. One of the problems with our business model when we built Strategy was that there very few people who needed enough experiment designs to keep us in business because they would get a design, take months or even years to run the experiment, then need our help or software for about two hours to interpret the results. When the trials are free and run this fast, the design of the experiment and the analysis become significantly more difficult and time-consuming. That’s fine with us because we have some unique advantages there.

The response to our efforts was the second surprise. We had never heard of this book, but it turns out to be very popular in the trading community. Many people have read, and those that haven’t, had at least heard of it. Many people had a common basis for understanding our work, and we are starting to have real conversations about applying this technology. We are moving on to models that we couldn’t even consider when the cost of trials was an issue.

Is Complexity a Bad Thing?

Complexity is taking a lot of beatings lately. It’s a carrier of misunderstanding, and we don’t want any misunderstandings – especially when we are full of fear. And there is no doubt that humans can very rapidly make things far more complex than they can understand and control. Our very capabilities carry the seeds of our own destruction. However, as long as you can understand and control complexity, it is your friend.

I’m no fan of complexity myself. I want everything not only simple enough so that I can understand it, but also so simple I can’t misunderstand it. I want it simple enough so that I can learn everything about it. Yet it is complexity that makes our modern life modern. Cars are more complex than walking, computers are more complex than pencil and paper, and a steel driving machine is more complex than a hammer.

A monkey understands a hammer. You could teach a monkey to use a steel-driving machine, but not to build or maintain one. Most of us are like that monkey. That is, we can only make use of a vast amount of the modern world because someone has found a way to package complexity in an easy-to-use wrapper. We can exploit complexity because someone else has taken care of all of the details. In fact, complexity beats simplicity on a wide variety of measures.

Once complexity gets out of control, it can be very dangerous. A drunk behind the wheel of a car and a madman with a nuclear weapon are just two such examples. Complexity is powerful, and control is easily lost. Simplicity may win in the long run, but complexity wins time after time in the short run. Are you going to bet on the man with the hammer or on the steel-driving machine?

I think we are all going to bet on the machine. It has too many advantages, each of which is an encapsulation of complexity. We want complexity on our side.

I think this is true when you are following or building an investment management system or a trading system as well. If you want to beat the other systems that are out there, you have to have more complexity on your side. You need to add the right kind of complexity, you need to keep it under control, and you need to deliver in an easy to use package even if you are the only user. The more details you need to keep in mind to use your system, the more you are likely to misuse it.

Let’s look at a simple investment plan. Buy real estate. Lots of people have made lots of money with a plan no more complex than this, but it is easy to see how a little complexity could make it better. For example, it gets considerably more complex if the plan is simply expanded to add one condition: “Buy real estate if you can afford it.” The added condition encapsulates a significant amount of complexity, and that complexity should make the plan work better.

If you actually want to make sure that you are getting a good deal, that complicates the plan even more. “Buy real estate when you can get it for less than it is worth for a price you can afford.” It still looks like a pretty simple plan, but we can all attest that our ability to state simply and plainly what needs to be done doesn’t really mean that it is simple or easy.

There isn’t really much simplicity around until we impose it by not thinking about most of what is going on. That’s not a problem for us. People are very good at ignoring most of what is going on. It’s how we get through every day.

A Thousand Trials

Usually, when you run an experiment, the primary cost of the experiment is the cost of running the trials. Designing the experiment is usually fairly simple, and never very expensive. The analysis requires some moderately specialized skills, but you can usually afford the analysis if you can afford the trials. Now, for the first time in our experience, we have discovered a field of investigation where the cost of the trials is almost nothing. This has profound side effects that we did not anticipate.

We should have because the most complex experiments were always run by people who had the least expensive trials, but even testing print cartridges makes a noticeable dent in a budget when you want to run a hundred trials. At the very least, the tests will still take hours, perhaps days, to run.

When we started applying Design of Experiments to financial engineering, we were strictly interested in intraday trading systems. Testing these systems had a moderate cost. For example, if we wanted to run sixteen trials on three years of data, it did not cost us much money, but it did take several months. That has a significant cost when you want to know the answers right now.

We wanted to talk about those experiments, but a number of factors prevented that We looked around for another system that we could experiment on. Ernest Chan recommended that we look at systems in Short Term Trading Strategies That Work and High Probability ETF Trading, two books by Larry Connors and Cesar Alvarez. As Ernie correctly perceived, Connors and Alvarez described systems that lent themselves very well to a Design of Experiments. We came up with lots of ideas for experiments right away, and we picked one out to try it out.

My partner Ron is very fast at this kind of thing, and he had a simulator ready to run the test after about a day of work. He had to organize the data because we had never worked with an end-of-day system before. Then he started running the test. We knew they would go quickly, but they only took about a quarter of a second to run a trial on fifteen years of data. It took far less than half a minute to run 56 trials.

This is as close to instant gratification as an experimenter will ever get! It took a few hours to analyze the results, but it was essentially “same day service”. You usually have months separating the planning of the experiment and the analysis.

Immediately, we began to think of more complex experiments looking at more factors and using higher order models. Since the higher order models required more trials, we rarely even thought in that direction. The cubic model was the highest order model supported by the software we were using, and we very rarely had the opportunity to use that. Most people are looking for experiment designs with less trials, not more. In fact, the point of most experiment design software on the market today is to get the most information out of the fewest number of trials.

Notice how the goal changes when the cost of the trials approaches zero. The goal in this environment is to get the best possible predictions without regard to the number of trials it requires. Another fifty trials? Another hundred trials? How about another thousand trials? We don’t care when the trials are free.

The Anatomy of a Quantified Trading Process

A quantified trading system is a system that can be automated, but it is not necessarily an automated system. A quantified process is precisely defined, uses numeric inputs, and produces measureable outputs.

Each quantified trading system starts with a Profit Plan. A profit plan is a theory that describes the actions that will be followed under the plan and explains how those actions will be converted into profit. Never forget that your profit plan is a theory. If it works, it is a good theory; if it doesn’t, it’s not. Theories of profit rarely become laws.

All theories start with observations. For example, you might first observe that stocks tend to bounce back after a sell-off. That makes you look closer, and you find this behavior seems to be a lot more prevalent in stocks that are doing well than in stocks in the general process of tanking. Your conceptual theory might then become “If a stock is doing well overall, you can make money buying on the dips Dump them as soon as the price bounces back.”

That’s a theory, but not a quantified theory. You can make it a quantified theory, as Connors and Alvarez did, by adopting numbers that define “doing well overall”, “the dips”, and “bounces back”. I am sure there are a lot of traders out there that did well enough with this rule with no quantification, but quantification brings two specific advantages. Once a theory becomes quantified, it can be executed by a computer, and once it can be executed by a computed, it can be tested exhaustively.

The process that Connors and Alvarez define on page 62 of Short Term Trading Strategies That Work is a good example of what is needed.


1. The S&P 500 Index is above the 200-day moving average

2. The 2-period RSI of the S&P 500 Index closes below 5.

3. Buy the S&P on the close.

4. Exit when the S&P closes above its 5-period moving average.

Step number one checks to see if the S&P is “doing well overall”. Is the 200-day moving average the best test of this? You can test for this. (We did, and found that every long-term average was almost equally effective so 200 is as good a number as any.)

Step number two is a quantified definition of how to spot a dip. You can now test whether this works, Is this the best indicator? You don’t know, but you can now test it against other alternatives. Quantification does not start with the assumption that you know the right numbers; it starts with the assumption that you can test and discover the right numbers.

Finally, step four quantifies what a bounce back is. Is the 5 period average the right indicator? Again, you can test for that.

A quantified system has rules that can be automated and tested. If you are automating a quantified trading system, keep in mind that the ability to run tests is likely to be the most important contribution of such a system.

High Probability ETF Trading: 7 Professional Strategies to Improve Your ETF Trading

If you are using any kind of a system based on technical analysis, you need to own and read this book as well as the earlier work by the same authors (Larry Connors and Cesar Alvarez), Short-Term Trading Strategies That Work. Both books are clear and understandable, and both lay out approaches that really do work.

The difference between the work of  Connors and Alvarez and almost everyone else is that they show you exactly how to APPLY commonly available statistical tools.  Everyone who has spent a couple of hours watching CNBC knows what the VIX is, for example, but it is just noise to the vast majority of people.  Connors and Alvarez provide explicit instructions that enable the application of this number to buying and selling decision making..

Take the 200 day moving average, for example.  A fine number, easily available from a number of sources, but specifically how should it affect your trading strategy?  Connors and Alvarez use this measure very simply and directly.  If the target security is above the 200 day moving average, only consider long moves.  If it is below that average, only consider short moves.

This is simple and measurably effective.  You can test any of the strategies in any of their books with and without this rule, and the difference is clear.  The strategies work when this rule is included; they don’t when it is not.

This book introduces their approach to short selling and money management tactics. The short selling tactics are a logical extension of their long strategies, but the money management is a new wrinkle. Like most of their presentation, it’s not fancy but it does offer a quantified approach for averaging into a position. The best aspect of everything that these books offer is a solid foundation for quantification because without that foundation, experimentation is useless and progress is uncertain.

The only quibble  that I have with either book is over the definition of the word “strategy’. As they say more than once in this book, it’s all about selling into overbought conditions and buying oversold securities. One strategy for which they present several effective tactics. I’m not knocking this. It’s a good strategy, and they make it better when they only buy in healthy markets and sell in overall weak markets.

They present a lot of useful statistics about a wide variety of indicators, but they are only looking for indications of overbought or oversold markets. The indicators they suggest are pretty reliable and work effectively when used as they suggest, but they are not different strategies. For example, you could not diversify by using several of their different strategies because they all tend to lead you into the same trades. That is, security A is overbought or oversold. Several indicators may reveal this, but that does not mean that there are several different opportunities.

A really good book makes you ask questions. For example: Which indicator is the best? Is there a combination of these indicators that is more effective than any single indicator on its own?

Connors and Alvarez also present a variety of indicators to tell you when to get out of a position. Each of these indicators point to the market no longer being oversold or overbought. Notice that the strategies are not looking to profit further in that direction. The end-of-day aspect of the exit rules may be trying to catch a little of that over-bounce. They make a point of saying that you shouldn’t get out during the day if you hit the target for that day. I suspect they are right, but that is one of the few assertions that they do not back up with statistics.

Engineering is all about gathering and evaluating measurements, then putting those measurements to work.  This book, like its predecessor, provides a good hands-on introduction to the basics of financial engineering.