Saturday, April 12, 2014

How to Deal With the Uncertainty of Trading

Reader Eldad Nahmany asks the excellent question, "What if I can eliminate uncertainty by accepting the true range of results that can arise from each trade?  I guess everybody said it before: accept the loss before you even enter the same way you accept the winner, then I can become a true observer of information.  The question now is how do I accept the loss before i get involved? I need to believe in the trade, no?"

It's a dilemma on the surface:  how do you maintain confidence and conviction in a trade and, at the same time, embrace the trade's uncertainty by planning its failure?

If you were a retailer and you had perfect knowledge of demand, you would always order the right amount of product in each category and never maintain excess inventory.

Once we accept that we cannot have perfect foreknowledge, then we accept the necessity of inventory.  We order extra product, because we don't know if customer interest will expand or contract; if a particular offering will fly off the shelves or languish there.

Of course, you could ask the retailer, "How could you believe in your product and yet still plan for inventory?  Aren't you preparing for your own failure?"

The answer, of course, is that any specific product might succeed or fail.  If you are a good buyer and merchandiser, you succeed across a variety of products.  You carry enough different products and price them well enough to ensure that you'll have some hits as well as some duds.  The duds will lose you a limited amount of money; the hits will be restocked again and again and will make up for the duds--and then some.

So it is with trades.  You don't need blind confidence in each position.  You need enough confidence in your overall trading ability to put on a variety of trades.  Some will be duds, and you'll catch those quickly and minimize their losses.  Others will be hits and you'll add to those as they prove themselves.  Any particular trade can be a failure; it's the edge across trades that makes you successful.

When people wax poetic about their conviction in trades, my emotional reaction is:  Whatever.  A trade is a bet at the poker table.  Some bets will work, some won't; some you'll size up, some you'll fold.  Whatever.  Over time, if you play the odds, you'll do OK.  Beyond that, it doesn't make a lot of sense to beat the chest and invite overconfidence bias to replace normal confidence.

Every forecast of a statistical model can be wrong.  Every trading judgment is fallible.  If you have a 50% hit rate on your trades and you trade once a day, on average you're going to have an occasion in which you lose every day for a week during a trading year.  That doesn't mean you're in a slump; it doesn't mean you should change what you do.  It's going to happen and you can mentally prepare yourself--and size yourself in such a way that five consecutive losing days won't take you out of the game.

The goal is not to eliminate losses--that would require omniscience.  Rather, the goal is to  anticipate losses so that you're never surprised, never overwhelmed, never thrown onto the back foot.  True confidence comes, not from believing that you must be right, but from knowing that you can survive and even thrive if you're dead wrong.

Further Reading:  The Power of Uncertainty   


David Ayer said...

Having an effective way to mentally frame your trades is one of the most crucial aspects of trading psychology. After entry any outcome is possible since conditions keep changing and other traders' decisions ultimately determine which outcome prevails. So always be uncertain and prepared for any of them.

Models provide us with differing probabilities for those outcomes. A loss is wrong only if the bet is against the most probable one. But a model that is 75% predictive (which many short-term retracement models are) still has a quite large 25% chance that price will not revert to the mean.

Another consideration is that different models, statistical or otherwise, can conflict. QQQ is oversold short-term, in a lower highs and lower lows short-to-medium-term downtrend, and finishing the building of the left shoulder and head of the medium-term pattern.

Market Monkey said...

I would add that the only way to remedy the apparent conflict between believing in the trade and fully accepting the possibility of loss is to re-frame what we are actually believing in. We need to believe in the process that we are using to take and manage trades, not the outcome.

It's often said that not-so-smart people talk about people, those of average intelligence talk about events and people who excel talk about ideas. We need to concentrate on being process-oriented (the idea) rather than outcome-oriented (the event). Then the outcome will take care of itself over the long haul...

Meditation is also an extremely powerful tool as it teaches us to not cling to, or reject, any possible outcome so that we are free to just observe things as they are rather than force our own agenda on to the situation (which always ends badly!)

Drew Coughlin said...

I would phrase this differently or extend the argument, "Every forecast of a statistical model can be wrong. Every trading judgment is fallible." While that may be true if you've built a forecast incorrectly (bad data, mathematical mistakes, etc.). I think it is much more important to recognize that the forecast and outcome is perfectly correct if the model, statistic, etc. is accurately constructed. It isn't that the forecast is wrong in this case, it is the fact that the unexpected, negative result is part of the model because all results are within the realm of possibilities. If I'm betting against an unlikely outcome, I have to fully accept and plan for the fact that the unlikely outcome is possible and can happen at any time. This is why we must obey stops and / or implement hedges.

Curtis said...

See my last comment on internalizing vs externalizing traders.

On uncertainty, information-processing (process based or i.p traders) traders embrace uncertainty

On "accepting the loss", i.p traders have a strong ability to know whether the trade will work out or not and try to take action accordingly.

The primary edge that an i.p trader has over a system trader is the ability to know when they are most likely to be right or wrong. Many system traders don't have that ability. This enables the discretionary trader the ability to dynamically control the leverage and the risk on the trade. By rapidly scaling the risk, increasing and decreasing, super normal returns can be achieved with only mediocre edge.

This contrast with the externalizing trader. The externalizing trader typically has a trading system that they will trade "blindly" trusting the results of the system. The externalizing trader doens't have to believe in the trade or not believe in the trade but what they do believe that they can't improve on the outcome via their discretionary judgement. You see how this contrast with the internalizing trader who derives their major edge from knowing if the trade will work or not. As well, some of the characteristics of system trading make it harder to improve those results through discretionary judgement. For example, most system trades are more selective, trade less frequently, and typically have a larger edge then the discretionary trader. So, there is a higher cost of trying to improve the results.

I'm not convinced that just anyone can become an i.p trader. With extensive research and development expertise, some system traders may still improve their results through another form of thinking we call systematic trading which involves if-then logic, case-based reasoning, and extrapolating or inducing hypothesis based on deep understanding of specific historical market behavior.

You'll notice Dr. Brett has moved back and forth over the degree, nature, and value of adaptability/flexibility in trading. Imagine for example, distinct "clusters" of ideas or concepts that when combined together that work. These clusters might be individual successful traders. On examining a cluster, one might choose to emphasize a certain characteristic of that cluster to teach a lesson. But, generalized lessons aren't of significant value because it is the "cluster" that produces the distinct advantage.

As an example, the cluster that the ip centric trader utilizes to advantage might be compromise of ability to generally predict the market under many conditions at a modest but variable rate better then chance and skill in using dynamic leverage. Under this sort of cluster, one could imagine many types of psychological issues and lessons that the trader may experience. This sort of trader may do well to cut losses quickly, add when confident, learn from mistakes, etc.

A completely different cluster would be analagous to the system trader who trades a system that can predict the market at a very high level under very specific conditions. As the trade has a high probability of working, the cluster might emphasize taking more risk in the trade then attempting to minimize losses. This sort of trader may be rightly encouraged to play the odds, trust the system, and to ignore 5 losing trades. "Can historical models adequately account for unique geopolitical events? Not without meaningful modification." --Notice, the system trader doesn't need to believe that their system accurately models the market under all conditions but only rather that over time the system models the market in a way that produces a profit.

I would say that a generalized "conviction" in trades isn't valuable but that the correlation between confidence and actual probability of a trade working is the hallmark, the defining factor of the successful i.p trader.

Matt Fitzgerald said...

I traded best (2010-2011) when I had very little knowledge of the market. As I learned more, I traded worse until my trades became predictably terrible. As I tried to trade against myself, I was unable to reserve this pattern. The common denominator was where I got my information (Yahoo Finance, Google Finance, established media sources). Once I subtracted them, I became very clear headed and a better predictor. I now only using thinkorswim charts (and I hastily avoid looking at any headlines). If it isn't a chart, I don't pay attention. This has helped dramatically as I wait for my next big trade....