Good morning all.
Day two of Steven’s guest slot and risk reward is the topic. A great post yesterday and I think todays is going to match it. If you missed it, day one is here.
Notional Risk-Reward and Expected Value
Welcome back, everyone. Today we put yesterday’s introductions and large-scale observations behind us and zoom in a little more to focus on a more realistic way to evaluate risk and reward. I should warn you, there is a tiny bit of math. But don’t worry it won’t kill you (or maybe it will, I really have no idea if math can do that).
I’m also going to try and keep this one shorter than the last, so please ask me whatever questions you might have in the comments below.
Risk-Reward is frequently expressed as something like 3:1. In other words, if you put on a long USD/JPY trade at 103.00, your stop is at 102.90 and your profit limit is at 103.30. In theory if you do that over and over and over again you’ll come out ahead if you’re right just a hair over 1/3 of the time.
Unfortunately that’s where many traders, both retail and doomed would-be professionals, stop the process. The thesis of this post is that the above ratio, sometimes “Notional Risk-Reward,” is crucial but by itself incomplete. We need to add the highly subjective and finicky concept on “expected value” to the pot.
So what’s expected value?
Let me give you two examples: first, roulette. An American roulette wheel has the numbers 1 thru 36, plus 0 and 00. There are 38 numbers total. If you wager a unit on one of these numbers, and you win, you get paid 35 times your original bet. In other words, your reward/risk ratio is 35-1. That might sound super, until you realize that the probability of winning (reward) is only 1/38, and the probability of losing (risk) is 37/38, which gives us an EV of:
{(37/38)*-1} + {(1/38)*35} = -5.26%
Expected value is the payout ratio plus your chance of winning the game.
The R-R ratio is meaningless without the probabilities. In the investing universe we can only be certain of one leg of the equation: the game’s payout ratio. You act as the casino when you set your stops and limits. You determine the game’s cost-payout ratio and every trade is a new game.
What we don’t know for certain is how many spots are on the wheel. In other words: how many different scenarios could play out once I put this trade on? To name just a few: earthquake, Abe makes a surprise hyperbolic comment about slaying the beast of inflation, a data surprise, a previously unknown merger announcement, ad infinitum.
A greater uncertainty means more spots on the wheel. And that means the less attractive a trade actually is unless you change your stop and profit limits.
Estimating how many spots are on the wheel of your trade is where subjectivity enters in. This is the art of trading. In large part, this estimation is what separates the heroes from zeroes.
Everyone here will have their own way of doing things, largely determined by the time frame they use. I work off Daily charts and zoom into the 4-hour’s to find a sweet entry point (5 and 15-minute scalpers, you guys are cowboys with much more courage than I possess).
My trades can last anywhere from 8 hours to several days. I like to take the 1hr, 4hr, 1d, 1w RSI’s and work them into a weighted average before the week begins. I do this for whatever currency pairs I am interested in as well as the largest single currency ETFs for all the major currencies. Doing this helps me put the strength and trend of each individual currency into perspective. From those ETFs I pick the most severely overbought and oversold currencies in order to see if I can find any clear instances of solid demand/supply or resistance/support.
Oftentimes the best play is to pit them against each other. Some of you may recall the spreadsheet John posted on my behalf two weeks ago suggested a pullback against the heavily overbought Swissie by the heavily oversold dollar. Granted, it took a week to play out, but it got there with my stop intact. The RSI’s mixed with my own forecasts of upcoming economic data and price action led me to determine the likelihood of this trade hitting was higher than 50% (I can’t remember the exact number – it was two whole weeks ago).
To recap: notional risk-reward should be combined with expected value to produce a more honest representation of the risk and reward of each trade.
Put more plainly: I do not think the Long USD/JPY trade above with the 3:1 ratio truly represents the risk you’re taking if the pair is in a strong and confirmed downtrend. You need to add in expected value.
Tomorrow we’ll begin evaluating some trades together. Three, actually: 1 from this week and 2 closed trades. And, due to popular demand, at least one of them will be a trade I lost a bunch of money on! See you then.
Contemplating risk/reward from Aleppo, Syria
