I used to have my desk in the trading room close to a team of analysts, this kind of modern oracles who appear on Bloomberg TV and predict the future of the currency pairs or anticipate the next move of the FED and I happen to know a bit about their job and daily tasks. An important part of their job is to be right: if they want to keep their position safely, they must be accurate in their forecasts most of the time. Our job, dark speculators and damned traders, is far from the endless Quest for the Truth of these analyst knights and totally materialistic : our job is to make money.
Fortunately for us, Fellow Traders, the need to be most of the time right to be a profitable trader is a Fallacy. As a good analyst can be a bad trader, a bad analyst can be a good one and my trading improved dramatically when I got that and I stopped focusing on being right but on making money. While in my trading having a view on the big picture and assuming some macro-economics hypothesis is of the highest importance (some profitable traders just don't care), I believe that one's trading shouldn't be driven by forecasts of where the market goes but driven by the events that occur on the way and success depends on the trader's reaction to those events. This said, in this post, I will focus more on the "need to be right most of the time" in order to succeed in trading.
A trader's performance is highly correlated to the Expectancy of his/her trading system, roughly the average profit or loss he/she can expect to make on one deal. The Expectancy can be estimated as follows:
Expectancy = avgProfit * %WIN - avgLoss * (100% - %WIN)
Where:
avgProfit is the average dollar amount earned in a winning trade
avgLoss is the average dollar amount lost in a losing trade
%WIN is the percentage of winning deals (the number of winners divided by the number of deals)
The rule is pretty simple: to be profitable in the long run, a trader must have a Positive Expectancy.
An example of negative Expectancy system comes from the Casino and the Roulette game. In the European Roulette, a winning straight up bet (on a single number) of $1 will pay you $35 (+$1 as your bet is returned) while your odd to win is of 1/37. This makes a slightly negative expectancy of 35*1/37-1*(100%-1/37)=-0.027 and believe it or not but the Casino only lives on these statistical 2.7% (the house edge) as in the long run and in average the players go to a sure loss. One can't beat the Casino at the Roulette in the long run as the game is structurally built to let the Casino profits from this house edge and the players can't change the odds against them. On another hand, I believe the market can be beaten, but to do so, one needs first to earn back at least the bid-offer and the brokerage fees that act as the market house edge, aren't the markets a giant Casino??? Obviously, the more you trade and the more you're leveraged, the more this house edge goes against your Expectancy.
Now an example of positive Expectancy, to stay in the gambling world, is card counting at Black Jack that lets the player know when the odds are in his/her favour and the expectancy is positive: it is forbidden by the Casino as counters would ultimately lead it to a certain ruin.
So trust your old friend Sauros: to be profitable in a sustainable way, keep your expectancy positive. Of course, it's easier to say than to get, the main issue is as we trade, we don't know what our own Expectancy is and particularly we don't know whether it is positive or negative. It can only be estimated, for instance using the formula above.
3 comments regarding the estimates of the Expectancy:
- The Expectancy can spend some time and a significant number of trades before it asserts itself, in the meantime, the estimates can be far below or above of the real Expectancy.
- Thus, assuming that your expectancy is positive, you need to survive until it asserts itself: you need to size your stakes in order to minimize the probability that you go broke before finally entering in the positive area.
- The estimates are computed and are valid under some market conditions and a specific environment and those change... For instance, some guys process, backtest and crunch zillions of historical datas and come up with set-ups that are supposed to work with a high probability based on the past. But when they ultimately trade those signals, it often appears that the success rate is much lower, shooting down the expectancy and they end up losing money as the market conditions have changed. In my experience of Market voodoo, it is wise to assume that the market conditions will go against the backtests at the moment you go live...
Now, let's come back to the estimate formula above, as you can see, there are 3 factors that can be worked in order to achieve a positive Expectancy: how much money you make with a winner, how much money you lose when you lose and how often you're right. The Holy Grail is to make a lot of money most of the trades and keep the losses small and rare, but there are traders using quite different combination of these factors and who manage (I guess) to be profitable:
"Cut the profits short and let the losses run" is the motto of many short term day traders and scalpers (even if they are not aware of it). The short term players need generally a high percentage of winning deals (high %WIN) to be profitable but to achieve this, they must take their profits very quickly and accept to take a bigger loss from time to time (AvgeLoss>AvgeProfit)...The argument is that in the short term, the markets are noisy and volatile, if you have a take profit closer to the spot than your stop loss (if any), the odds are in favor that the take profit is hit first. Those players generally add on some technical analysis or trend following algorithms in order to increase the probability that the position wins.
Such trading systems can only fly if the rare losses don't erase the multiple locked profits and keep overall... the expectancy positive (I know I'm repeating myself)
The low risk-high reward players try to optimize their expectancy in trading some set-ups with both favorable odds of success and losses kept lower than profits. The stop loss is generally closer to the spot than the take profit (avgLoss<avgProfit) and positioned in a way it' hit only if the price breaks some obstacles (supports, resistances, etc) that are assumed to hold. One will notice here that in the case the average loss is below the average profit, it's enough to be right only half of the time to have a positive expectancy.
The traders who "cut the losses short and let the profits run" are generally long term players using a low leverage. By applying this principle, those players obviously have a higher average Profit than their average Loss (avgProfit>avgLoss) but tend to show automatically a lower percentage of winning deals. This is very difficult to trade because you need to have a strong faith in your trading system as a long losing streak will inevitably erode your confidence in it. Once the confidence is hurt, you will tend to take your profit earlier as you let them run the previous times they just vanished ending up with a loss. Ultimately taling your profit earlier will decrease your Expectancy. Another danger is the "break-even effect": after a long losing streak, you tend to cut the position when you come back to the break even after a long journey in negative areas near hell. Hey remember, your job is not to make back your initial losses and end up at zero but to make money... I would say that most of the "legendary" traders lie in that category and I would say that I guess they are right only 30-40% of the times. As we discussed in my previous post, the trick is when they are right, they make it big!
Personally, I've tried all those types of trading (and much more!) and now I apply in my trading mainly a combination of the last 2 principles. Frankly, I happen to wonder from time to time (particularly when I'm in a looong losing streak) if I wouldn't be a happier trader if I was most of the time right rather than spending my time to be wrong and to wait for THE deal that will make me finally profitable. Well, that's not the point here, the thing I wanted to show here is that you don't need to be right to be profitable and I hope I managed to convince you my fellow Traders.
----------------------------------
I've posted some additional thoughts about this post in the Hand of Scalpuman, the traders' forum of TLofT. Please feel free to come and comment this post and participate to our market discussions.
Posts
