Kelly Criterion & Risk Of Ruin As Risk Management Tool

[bpr]

I was correct on that.
Note I said this is a bottom|top catching strategy,so the distance it runs(representing profit), is 3 to 6 times the distance between where the signal occurs and stop(representing the loss incurred).
A signal at the same direction as major trend runs longer in distance than a signal against major trend.

This is also one of the reasons I dare to risk 50% of my capital.Because I profit 3 to 6 times more than my loss.

very hard to beilve a 90% win system with risk to reward 1:6
u should be multibillionair already ??

your system kelly is 88% half kelly 44%

with 90% win
2 consecutive loosers probability is 100%
3 consecutive loosers probability 10% approx
4 consecutive loosers probability 1% approx
5 consecutive loosers probability 0.1% approx

So u should take that into account instead of blindily going with half kelly

20% per trade risk would be more appropriate.

 

[bbpp]

very hard to beilve a 90% win system with risk to reward 1:6
u should be multibillionair already ??

your system kelly is 88% half kelly 44%

with 90% win
2 consecutive loosers probability is 100%
3 consecutive loosers probability 10% approx
4 consecutive loosers probability 1% approx
5 consecutive loosers probability 0.1% approx

So u should take that into account instead of blindily going with half kelly

20% per trade risk would be more appropriate.

As I understand you already consider the factor my win is 3 to 6 times of my loss?
The strategy was built during the past year and I did win some money but lost on other strategies I have been testing.As I am not sure it is the best strategy I can get.

 

[bbpp]

I had bottom signal for 10 years note on 7/29/2019;
top signal for euro on 6/25/2019;
Bottom signal for euro on 7/25/2019.
The last signal failed.

See for yourself on chart how did they work.

For those who are interested, I will post my real time signals in following thread.

https://www.elitetrader.com/et/threads/position-trading-calls-and-ideas.335664/

 

[ElectricSavant]

This is my opinion and you need to find what works for you.

I find that making decisions on trade size should be based upon the system you are trading. You should rank the "trade-strength" based upon the system not some sort of gambling scheme.

Again...The amount that you trade should directly correlate to your signal used.

Es

 

[ironchef]

It was a typo.
I meant stop hit 1 out of 10.
Thanks for correction.

I guess putting 50% of my capital at risk is roughly equal to 1|2 Kelly but am not sure.

 

[777]

From my initial tests it looks like this is not worth pursuing. Usually, for most systems I tested the lower the kelly fraction the better that ratio is. So if someone wanted to maximize for that, they would just bet a really tiny amount that wouldn't be worth if at all. In general it seems that the closer to the full kelly one gets, the more capital growth vs max DD is maximized
(which can be measured by things like the MAR ratio). This leads to other risk adjusted metrics (like Sharpe, K-Ratio or that CAGR/Median DD ratio) look worse relative to smaller fractions. But if someone wants to maximize for those metrics, than they might just keep decreasing their bets to the point its a waste of time. So in the end its all about a balance that one has to find

Few successful risk takers use full Kelly, though they often use fractional Kelly.

In simulations, Kelly makes capital grow at the fastest possible rate with no chance of going broke. Consistently using more than 2x Kelly guarantees you lose money even though you have an edge.

Full Kelly runs into many problems and here are a few:

You may be wrong about your edge.

You may be unsure about your edge.

You can not keep reducing your bets/trades when losing because of opportunity cost or other factors.

Expenses are the same as loses.

The downside swings may not be suitable for temperament, especially as one becomes less sure of edge.

Unexpected events such as loss of capital in system failures, fraud, cheating, etc, can be devestating.

Sometimes people or groups have too much money to deploy full Kelly: blackjack table limits, limited number of shares at the desired price, etc.

 

[ironchef]

Few successful risk takers use full Kelly, though they often use fractional Kelly.

In simulations, Kelly makes capital grow at the fastest possible rate with no chance of going broke. Consistently using more than 2x Kelly guarantees you lose money even though you have an edge.

Full Kelly runs into many problems and here are a few:

You may be wrong about your edge.

You may be unsure about your edge.

You can not keep reducing your bets/trades when losing because of opportunity cost or other factors.

Expenses are the same as loses.

The downside swings may not be suitable for temperament, especially as one becomes less sure of edge.

Unexpected events such as loss of capital in system failures, fraud, cheating, etc, can be devestating.

Sometimes people or groups have too much money to deploy full Kelly: blackjack table limits, limited number of shares at the desired price, etc.

Great comments.

What I found was the uncertainty of my positive expectancy affected the Kelly #. Mine changed from year to year and therefore there was an uncertainty band around Kelly. Now I don't calculate Kelly when trading, just intuitively use a bet size of fractional Kelly.

 

[chrisbamtrader123]

I thought according to Ralph Vince that Kelly was no good for trading when there were always different out comes possible. I believe he was adamant the optimal f was far superior for trading. Does anyone have any thoughts on this? Thanks!

 

[ironchef]

I thought according to Ralph Vince that Kelly was no good for trading when there were always different out comes possible. I believe he was adamant the optimal f was far superior for trading. Does anyone have any thoughts on this? Thanks!

Trading is all about probability. A single trade can have different possible outcome, like a hand of blackjack or poker, governed by chance and randomness. But with enough trades and a large sample size, your method should produce an average win rate and R:R from which you can certainly compute your Kelly. Though unlike blackjack and poker, the expectancy may change over time so you have to update your Kelly periodically.

I am not a professional and may very well be all wrong. You should consult a professional.

 

[chrisbamtrader123]