stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

How to Trade Brilliantly but Still Loose
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=160		 Page 1 of 1
Author:  garyfritz [ Fri Dec 16, 2011 3:03 pm ]
Post subject:  Re: How to Trade Brilliantly but Still Loose
Optimal-f and Kelly calculations tell the "optimal" leverage to maximize profits. What many people don't realize is that even if you don't do a Corzine, you still risk losing if you overleverage. Any risk level under OR OVER the optimal point reduces your profits. If you risk more than the optimal fraction, you make less profit than at the optimal level, up until the point where you go broke.

Extreme example: let's say you're playing a game where you flip a coin. Heads you win $100, tails you lose $1. What great odds! What a great trade! So you over-leverage and bet 100% of your account. And then the coin comes up tails...

I used to race Porsches. Racing is an interesting analogy to trading, with speed == leverage. The faster you drive, the faster you get to the finish line... unless you drive TOO fast. Drive TOO fast and you end up crashed out in the weeds. Leverage TOO high and the same thing happens.
Author:  garyfritz [ Fri Dec 16, 2011 5:31 pm ]
Post subject:  Re: How to Trade Brilliantly but Still Loose
I agree, Khalid, but I take it farther than that.

Betting anywhere near optimal is stupid. Even if you think you're prepared for the 90%+ drawdowns you can (and almost certainly will) experience trading at "optimal," things can change. What was "optimal" during your backtest history may turn out to be wildly over-optimal in "unexpected extremes." Trading at "optimal" is gut-wrenching at best, and suicidal at worst.

Trading well below optimal is the sensible approach. That reduces your drawdowns and reduces your risk in unexpected conditions.

The key is analyzing your trade history and determining what is safe FOR THAT SYSTEM. 1% is not a magic "good" value. There are plenty of trading approaches for which 1% is too much. There are many for which 1% is far below a safe level. If you have a good system with a high "optimal" value and you're willing to accept far less profits than the system can safely deliver, then that's great, trade 1%. You'll take very little risk and you'll sleep well. If you want better return for your money and time, analyzing the system's behavior and determining a more-aggressive but still safe risk level can greatly increase your returns.

Running $400M to $200M is really only exceptional because of the size of the loss. A 50% drawdown is "no big deal" for someone betting aggressively. The returns you can get at the higher leverage are worth the risks. Which would you rather take, a 50% drawdown from $1M or a 10% drawdown from $100k? Lower risk might keep your drawdowns small but limit your growth to $100k. Higher risk might subject you to higher drawdowns but build the account so much faster so that even the deeper drawdowns are still ahead of the lower-risk scenario. The right choice for any particular trader depends on your experience, your risk tolerance, and your confidence in your risk analysis.

Example: in one of my backtests I start with $10k and trade the system at various fractions of the Kelly value. I can trade it at 5% of Kelly and keep my worst-case drawdowns below 8%. After 18 months the account grows to $22205, then it goes into a drawdown. The 8% DD takes it down to $20522. Now if I trade it at 20% of Kelly, my drawdowns grow to almost 30%. But that same drawdown took the account from $196532 to $139232. In the deepest depth of the worst drawdown, the more-aggressive account had over 6x more profit than the conservative low-drawdown model. (You don't want to hear the returns I'd get if I was willing to stomach 50% drawdowns. :lol:) To me, that's worth the additional risk. To someone else, it might not be.

(Side note: in this backtest I calculate historical Kelly values once a quarter, and trade forward with those values for the next quarter. In the backtest, it turns out the "optimal" value is not 100% of the calculated Kelly, but actually around 79%. That's a real-life example of the "optimal" value on the backtest being over-optimal in forward trading. In this example, trading at 100% of Kelly makes as much as trading at 55% of Kelly, but the drawdowns at 100% are much worse.)

Bottom line: NOBODY should take high risks just because he thinks he'll make more. Anybody who can't or won't carefully analyze his trade history is trading half-blind, so he should keep his trade risk low. (Just like you should drive really slowly if you can't see through all the mud on your windscreen.) 1% is a reasonable "guess" assuming you have a clearly profitable system.

I prefer to clean the mud off my windscreen and determine a safe speed to drive. :D
Author:  garyfritz [ Sat Dec 17, 2011 1:40 am ]
Post subject:  Re: How to Trade Brilliantly but Still Loose
I understand your point, Khalid, and I would love to be in a situation where I can comfortably risk only 1% on my trades. But for the capital I have to trade, and the systems I have to work with, 1% risk produces too little profit for it to be worthwhile, to me. I'd rather be a bit more aggressive, after carefully studyinig and understanding the characteristics of the system, and go for higher returns.

To me it makes little sense to choose an arbitrary number like 1% just because somebody else uses it, no matter how much money he made. That's like saying "I'll never go more than 50 kph, whether I'm driving a Ferrari or a bicycle, because I believe Lance Armstrong never biked faster than 50kph." For the bicycle, 50kph is a sensible upper limit. For the Ferrari it's silly. With trading as with driving, I think you should adapt your "speed limit" to the vehicle you're driving.

(And yes, I know Lance has biked a lot faster than 50kph. Guess what, I'll guarantee almost every "super trader" also traded riskier than 1% in their careers, before they got super-successful and became more conservative.)
In Kelly's original paper, the criterion is only valid when the investment was played many times over, with the same probability of winning or losing each time, and the same payout ratio. I cannot for the life of me understand how one can apply such in real life trading.
There are enhancements of basic Kelly that account for real-life trading by using the variance of your trades. In my case I don't use that, because in my testing of my system I find that it doesn't make a significant difference. You don't have to calculate your risk out to 6 decimals. I really don't care if the results are off by 5% or 10%. That's probably much less uncertainty than what's presented by the market itself.

I've tested several years of trading, including almost a thousand trades. They are not identical, like e.g. rolling dice would be. But they are statistically similar enough that I'm very comfortable with the result. Especially since I've forward-tested it with varying Kelly fractions and, even though the Kelly calculations aren't 100% accurate for varying-size trades, I like what I see. I'm in the process of expanding that to a full 5-year walk-forward optimization analysis, which will very closely mirror how I trade in realtime. If that continues to confirm the results I've seen already, I'll be VERY comfortable trading it with my more aggressive risk profile.

The biggest risk will be that the market changes and breaks the system. That risk is present in any system, no matter what level of risk you take. If that happens, my losses will be bigger than they would have been at 1%. But most likely my winnings will have been bigger too, so I'll come out ahead. I'm very comfortable with that. Other people might not be.
Author:  garyfritz [ Sat Dec 17, 2011 1:49 am ]
Post subject:  Re: How to Trade Brilliantly but Still Loose
Let me add a reinforcement of Khalid's comments in case somebody misses the point in my wordy dissertations:

Trading at 1% is an accepted and relatively low-risk way to trade. If you don't know exactly what you're doing, and if you haven't carefully tested and characterized your trading, you have no business trading at a higher and riskier level.

Only if you have characterized your trading results, and are confident that they accurately reflect what you will do in the future, AND only if you fully understand the ramifications and likely outcomes of using higher leverage, should you consider using more-aggressive position sizing.
Author:  Aussiedoc [ Mon Jan 02, 2012 8:01 am ]
Post subject:  Re: How to Trade Brilliantly but Still Loose
Garyfritz, Khalid et al. With your permission I would like to chuck my two pennies in the hat - from a totally different, yet similar I believe, perspective.

As a recent newcomer to world of financial trading; my knowledge of it and what motivates people trading within it is petty much non-existent. As a medical practitioner this past 40 odd years, reading through this thread I noticed a similarity (maybe not expressed specifically as such) to the fact that all, most everyone I have met so far has really wanted to know, is 3 things:- 1. Can you fix my problem? 2. How quickly? 3. How much?

One of the things I learned about people, very early in the piece, is that, similar to trading; we also have mental, physical, spiritual, social, financial and emotional Support and Resistance points - controlled by our internal personal belief system. Support is what you know you know/what you know you don't know. Resistance is Fear of the unknown!...What you don't know you don't know! E.G. Do something you have never done before - Race a car! Cook a meal for your new parents-in-law! Do an 'instruments only' night-landing on a strange airfield! Scuba dive - solo! Become a parent! Trade for the first time! You will very quickly find your Support and Resistance Levels.

All knowledge can be divided into 3 areas. 1. What you know you know. 2. What you know you don't know! 3. What you don't know you don't know!! Everything anyone has ever learned in their life has come from area 3 (what they didn't know they didn’t know) - until some 'AHA/Light Bulb" circumstance or event occurred that that added to their knowledge/belief base.

Like most kids my age learning to drive in the 40s; I cringed every time I crashed the manual shift/double clutch gearbox (with my Dad sitting alongside me) and ''Kangaroo Hopped" every vehicle I drove until I knew the "How to properly change gears in a car/truck/tractor rules" off by heart.

Like everything in Life, the formula for Belief is simple - practise the rules until you don't have to practise the rules. The great difficulty we all have is ego - believing we are better than the rules. This is where discipline comes in. Most everyone says the want to be rich. They buy a ticket in a lottery so they can win heaps of money instantly - and lose it almost as fast because their internal belief system is still stuck somewhere between "Welfare & Middle Class."

Cynical?? Yep!! Mother Nature is a great teacher. We get sick/better - supported by our internal belief system - 1 cell x exponentially - until the condition reaches a critical mass. Our finances do the same! I was speechless when I was first shown the leveraging power of a small amount over time.

What makes threads like this so fantastic is that you, and everyone else involved are working very hard at shifting your internal belief systems WAY above 'Average' - by being prepared to put your egos aside; 'stick your heads on the block'; and dive into area 3 by playing volley ball with new/challenging ideas for 3 or 4 sets - and allowing others to share in that knowledge (The Piece Of Jade) to whatever extent we are able..

Boys, I take my hat off to you all!

Thank you all very much! I haven't had so much fun - and learned SO much I didn’t know I didn't know - in years. Wendy (reading this over my shoulder) says thanks!! She reckons it's taken 10-15 years off - mentally. Wants to know if you can do anything for me physically?? :roll:

Thanks now! Take care.
All times are UTC Page 1 of 1

Powered by phpBB® Forum Software © phpBB Limited