Optimal Leverage in Futures
- Thread starter Intutive
- Start date
Is it? 
. What's in a nick anyway?
It seems to me that the ability to ride out the noise is a function of leverage / money management. That said to not play like a pig when the ducks are lined up would be suboptimal. Where and how to draw the line between minimizing tail risk drawdowns and maximizing gains?
. What's in a nick anyway?It seems to me that the ability to ride out the noise is a function of leverage / money management. That said to not play like a pig when the ducks are lined up would be suboptimal. Where and how to draw the line between minimizing tail risk drawdowns and maximizing gains?
I use a formula based on drawdown and maximum loss, as well as volatility and a safety margin to come up with what I think is the best amount of capital to use per contract. Then I multiply the amount of contracts up to a level where the maximum loss/drawdown of each contract would equal an account drawdown that I'm comfortable with. The leverage then really just depends on that, I don't care if it's 1,5 or 5 as long as the account drawdown is considered.
I think that finding the optimal relation between leverage and account drawdown is important in order to optimize returns on a strategy. A strategy that returns 10% per year may well be better than a strategy that returns 30% (per one contract) if you're able to leverage it more because of a lower drawdown/max loss.
I think that finding the optimal relation between leverage and account drawdown is important in order to optimize returns on a strategy. A strategy that returns 10% per year may well be better than a strategy that returns 30% (per one contract) if you're able to leverage it more because of a lower drawdown/max loss.
Thank you both.
You raise an important point and I've noticed many people do not differentiate between favourable and unfavourable volatility. If you are long, only negative volatility is bad and vice versa (well even that is sometimes an opportunity if you can average down with nothing having materially changed in your call).
Positive volatility is always good i think (except in situations where it breeds overconfidence and therefore increased risk taking).
A case in point is investment in house or property. By and far, a majority of people make money in property - a levered investment (4-5x or as much as 10x). The reason they do is because 1. Negative volatility in that asset class manifests less often (say 10-20% of the time) and 2. Sitting on it for a long time. Absent these two conditions, the returns would be worse.
Similarly if an underlying security/trade exhibits lower historical negative volatility - one ought to lever it up more. That said, historical volatility does not capture unknown unknowns (where tail risk comes in). So there should still be an outer bound to leverage based on max. tolerable drawdown.
If your expected value for a trade is significantly positive (risk/reward very asymmetric), kelly correctly postulates you bet more. But kelly doesn't account for leverage as far as I know it.
You raise an important point and I've noticed many people do not differentiate between favourable and unfavourable volatility. If you are long, only negative volatility is bad and vice versa (well even that is sometimes an opportunity if you can average down with nothing having materially changed in your call).
Positive volatility is always good i think (except in situations where it breeds overconfidence and therefore increased risk taking).
A case in point is investment in house or property. By and far, a majority of people make money in property - a levered investment (4-5x or as much as 10x). The reason they do is because 1. Negative volatility in that asset class manifests less often (say 10-20% of the time) and 2. Sitting on it for a long time. Absent these two conditions, the returns would be worse.
Similarly if an underlying security/trade exhibits lower historical negative volatility - one ought to lever it up more. That said, historical volatility does not capture unknown unknowns (where tail risk comes in). So there should still be an outer bound to leverage based on max. tolerable drawdown.
If your expected value for a trade is significantly positive (risk/reward very asymmetric), kelly correctly postulates you bet more. But kelly doesn't account for leverage as far as I know it.