Hey do you mind explaining why you are doing this?
My idea is to build a strategy that deliver a small but steady income (Initial Reward) day over day. But as you said, it assumes more than could be experimentally expected. Such as low costs, infinite underlying's daily range, infinite capital, and infinite run time per day.
However, the value is in the eyes of the beholder.
This appears to be a type of reverse Martingale type of betting/risk if I understand it correctly. You appear to be increasing the required minimum REWARD before taking on the next risk in the next trade.
You got it perfectly. I thank you for you constructive and insightful reply.
That does not make sense to me because I don't know about you but my typical Risk:Reward is 1:2.5. Sometimes I get 1:4 or even 1:6 but usually about 1:3 or 1:2.
I began a journal not so long ago. So it's still in its infancy and I've got no internet since a week ... but, even if rare, I achieved few 1:10 risk to reward. On the ES, with a 1 point stop. The tighter the stop, the easier it is too maximize that ratio. However my average ratio is much lower (around 1:2).
I try to take trades with higher rewards obviously but if they were readily available I would take them every time. Your 1st post Implies you have those in abundance. I don't. My REWARD ratio is calculated based on my Target price but I don't reach my target every time. So even though I may think max reward is 4:1 I may get stopped out or have an exit signal at 2.5:1.
I totally agree. Why bother with small rewards if bigger ones are around the corner. However the actual reward is a know unknown. As you pointed it out. We expect a payoff with related frequencies for each and every outcome. We can take a profit while the market actually went farther in our favor. But inversely, we can expect more from the market where he actually reverse toward our entry price ... My point is that we don't know what is readily available. At the start of the day the daily range, hence your ratio, is to be built over time. Who knows ? Of course it is bounded. One can't expect unreasonably from the market in one day. Relative to it's historical behavior, what's in play, and the expectation of the agents. The risk is Known. But the reward is unknown.
Also, why increase the reward? If you are worried about losing too much and want to assure you have a win that puts you in positive PnL then lower your risk each trade rather than raise the reward.
I am not worried to lose too much. I am expecting to make 2 points a day. By offsetting the losses, as they accumulate, until the reward is hit. Your suggestion looks like a variant of this system. And at the end you get less than the initial reward (The main point this trick is build for). And it implies continuous divisibility of the risk. What if you can trade only one lot ? You can't lose less than 1 tick.
If you change your reward ratio every time you are in essence trading 4-5 different systems because the R:R is different every time. If you flip a coin and tails you lose and heads you win then every 1:2 trade would mean if it lands on heads you win $2 and tails lose $1. But if the outcome of tails means that next time you get $3 (1:3) or even $7 (1:7) if you get heads then why ever even play the 1:2 game?? These high reward ratios are 3 different systems. If they all had the same probability of winning play the highest reward one every time. Doesn't make sense to EVER play 1:2 game in that situation.
I don't agree that it implies different strategy. One could set different risk:reward conditionally adjusted to some metrics. But it's true that the ratio is set by the strategy. So here we have to set the ratio independently of the setup. Let the trick do it for us. To be honest, at the beginning I was thinking about generating semi-random signals. Or implement that strategy into a mean reversion, counter trend fading, system. Got to test which bias is best.
What is your ultimate goal? To maximize profit and minimize risk, right? Then it seems to me (it's been described plenty of times before) you should risk a set amount and back test and forward test your system. THEN after forward testing for several hundred trades then you have a REAL expectancy of your system. Now you play with optimizing your risk based on how much drawdown you are willing to tolerate and maximize returns with less drawdowns. This is done with a Montecarlo style of analysis on your system based on your expectancy and total number of trades.
I can't agree more. Yes. Yes & Yes.
Thanks for your inputs.
Really appreciate.
