Friday, 29 January 2010

More Trading Opportunities or Surer Trades - Which is Best

In talking with Kiki during our tweaking the FloBot's win rate, we had a discussion of whether it was better to have more trading opportunities or whether the trader should be more confident that the trade he was putting on was a winner.

I can make it pretty sure that the trade will be profitable, but then I will have very few trades. It's like having a slider between % win rate and profit per contract per week.

Let's say that a trader needs to make $500 per day trading.

Trader A can have a trading plan that says that he will only take trades that are 90% sure to make 1 point with a stop loss of 2 points and he will trade mechanically. That means that a trade will only make $46 per contract per trade. On back testing, Trader A finds that there were on average 10 trades a week of which 9 made $46 and 1 that lost $104, a net $310 per contract per week.

$500 earnings per day is $2,500 per week. Therefore to meet his goal, Trader A would trade 9 contracts per trade. His risk per trade is 9 times $104 or $936.

Now, let's look at my methodology. I'm Trader B.

I have about a 75% win rate on inside out trades. I trade in multiples of 3 contracts. My risk per trade (per every 3 contracts) is $312. My average net per day is over $900 per 3 contracts or $4500 per week.

So Trader A's Risk/Reward is 2.67 to 1

Trader B's Risk/Reward is 14.42 to 1.

Most traders focus their energies on deciding whether to take a trade or not even when it exactly meets their trading plan's rules. Why? To try and filter out a loser. For me, trying to be surer means less profit and really more risk - the risk of going broke as the statistics of the 90% of traders who do not survive, demonstrates.

On those figures it looks like being surer that a trade will be a winner is not the way to make money. As Pete said, "Get the trade on and manage it."

Looking at trade statistics is important to me as it helps me make decisions. Using a program like MSA from lets me do a lot of "what if's" when I want to tune a setup.


  1. Hi,
    I read your blog very day. I appreciate what you have done to help others.
    Here, just my calculation. I hope it will be helpful.
    Trader A has 90% win rate. So Trader A's expected profit per trade is 46*0.9=41.4 and expected loss is 104*0.1=10.4. Then, Trader A's Risk/Reward is 3.98.
    Trader B has 75% win rate. If his average net per day is over $900 per 3 contracts(assume only one trade), we can infer that his expected return per day is over 900. His expected loss per day is 312*0.25=78. Then Trader B's Risk/Reward is over 900/78=11.54. Great!
    Thank you again.

  2. A little confused about this calculation. Are you assuming that you will pull 6 points per contract on every three contracts traded? That is pretty ambitious unless I am misinterpreting what you are saying.

  3. Sorry,
    I figured out that I made a mistake.
    Actually, if Trader B's average net per day is over $900 per 3 contracts, his expected profit per 3 contracts should be over $978,average return plus expected loss $78.
    So, Trader B's Risk/Reward is over 12.54.
    Hope this time I am right.
    Good luck for you and Kiki.

  4. Anon, you asked: "Are you assuming that you will pull 6 points per contract on every three contracts". Not exactly, they were the recent stats of inside out trading. What I am saying that I believe you make less profit if you try and be too sure. Being "discretionary" allows you to take what the market gives.

    Annons, you can look at these stats in a million ways. I interested in a lot more than the win rate. This is just one example.

  5. Well said EL, kinda like today. Slapping on a trade below 77 or (holding one from above) gave 1070 as a not so ambitious target given that the market was literally sitting on stilts with the gap acting as a magnet from below.

  6. EL
    very nice blog .I think I have not problem with context .I spent a lot time study your charts and try find out feeling and nuances for setup entry and exit orders. I have to admit that I do poorly with it.Anyway I would like ask you.Assume that you splited yesterday profile for two distributions and you have two value areas now. Today price leaves one VA and entry to other. IF PRICE IS ACCEPTED ,DO YOU EXPECT THAT GOES THROUGH WHOLE VALUE AREA ??. So if entry through LVA and is accepted in VA ,does reach HVA ??? Thank you

  7. Pavel, if you have a correct split then the same "rules" apply as to a single distribution day. Its distributions that count, not days. So yas, once price is accepted inside a previous days overlapping VA, then the expectation is that price should go to the other side of value. However, it depends on the momentum and order flow. Often, if price has already gone through one VA and enters the second profile then the second POC may stop price, especially if it is a very wide VA.