Matthew,
The majority of systems posted on C2 employ some degree of leverage, implying an expense for margin interest. I know C2 stats calculate commissions (and by the way, how much are "typical commisions?") but do the stats also include margin interest? If so, at what rate?
Thanks,
Daryl
[LINKSYSTEM_26717594]
I know the question has been asked before, and the answer is no. But as a caveat to that, the higher than normal commission probably makes up for it.
Beau,
Thanks for the answer. Could you tell me where the commission rates are listed?
Thanks,
Daryl
[LINKSYSTEM_26717594]
Great point about margin rates. In the best case scenario, we could all plug-in our own settings for our own brokerages and have C2 subtract values from there. Creating ways to compare systems on a post-fee basis would be awesome. Margin rates are another variable that would be necessary to subtract from the pre-fee returns to make an apples-to-apples comparison between systems.
Post-fee returns = pre-fee returns - commissions - membership fees - margin fees
Absolutely, TJ. Because comparing equivalent systems that differ only in their use of leverage (2:1 to 4:1 for example) will have big differences in returns and risk. Comparing non-equivalent systems (which applies to the systems posted on C2) becomes exponentially more difficult making your methodology a great suggestion.
I hope Matthew gets a chance to read and consider this…
Sincerely,
Daryl
[LINKSYSTEM_26717594]
Matt said they were based on IB commissions.
Daryl. It doesn’t matter if you’re talking about 1:1, 2:1 or 4:1. Just scale it down by taking powers to the 1/2 or 1/4 to account for compounding. Then you’ll get the 1:1 rates. This isn’t a really big issue.
Say APR at 50% at 2:1 becomes 1.5^(0.5) or 1.2247449-1=APR% of 22.47. Do that for Drawdown as well and there you are.
It’s just like what I explained to you when you asked about doubling your position size what the effects would be.
Right, understood. But what was brought up earlier is that these calculations should be reflected in the C2 stats. Otherwise, the stats are reflecting unrealistic performance. For example, a C2 system that trades 2:1 as opposed to a system that trades 12:1 is currently paying the same amount in margin interest (nothing). The advantage the system using the lower leverage would normally enjoy is going to be marginalized, while the high leverage system is going to be substantially benefited by the absence of this cost. I understand the mechanics of it, just not why this potentially large effect on performance is not accounted for…
Sincerely,
Daryl
[LINKSYSTEM_26717594]
I wouldn’t worry about it. Much beyond 5:1 will blow up, and that’ll be reflected in the drawdown stat.
Beau,
What you are saying is very interesting and opens up some questions I have had in the back of my mind for some time. I have seen numerous highly leveraged systems fail after varying times of success, but I never thought of it as a correlation between high amounts of leverage and failure. My assumption has always been that a system can be highly leveraged as long as the corresponding hedge is (nearly) equally leveraged and highly uncorrelated. Is this inaccurate? Could you explain this in a bit more detail when you get a chance?
Thanks.
Sincerely,
Daryl
[LINKSYSTEM_26717594]
It’s about the position sizing options. In Wealth-Lab, there are certain ways to control drawdown and risk in a system. The trades you take and your position size will control your decisions.
My preferred method is a percentage of equity that will fluctuate with the size of my capital. This makes most percentage moves their equivalent to your profits.
The other way, that has less drawdown, generally, is by a fixed position size. This usually has less drawdown in the long run, because after big wins, you don’t use the excess to invest in other trades. You probably could just keep it, if that was your plan.
I can tell you from experience out of the hundreds of systems I’ve written that not one system could survive much beyond 5:1. You’d quit after the first 30% haircut. I can guarantee that, basically, because I understand the psychology of it. I could do it through options, but the drawdown would go to 30%, and that’s too much.
On this:
I have seen numerous highly leveraged systems fail after varying times of success, but I never thought of it as a correlation between high amounts of leverage and failure. My assumption has always been that a system can be highly leveraged as long as the corresponding hedge is (nearly) equally leveraged and highly uncorrelated.
The reason they fail is an overreaching amount of APR. There’s very few systems that have 200+% APR’s with less than 20% Drawdown. Maybe someone could point me out one. Risking a third of capital requires a 50% return to overcome, and more than that if you’re risking 45%, as some systems on here do. The insurmountable task of getting back to your highs affects your trading psychology, and you get nervous about the next several trades. What if it gets worse? You’ll be down on every trade you take at some point, but most people only think of the fact that they could be up.
I don’t think your hedge is what is important, as much as your timing is. Benjamin Graham, what to buy is probably less important than when to buy, just to paraphrase his quotes out of Securities Analysis. I think you can understand that. So even if I were to hedge my stocks to mitigate my leverage, the flip side is I could just take a smaller position if I was worried about it. Too many vendors have not fully understood the extent of the leverage they are taking on through futures contracts and especially forex. I’ve never been one to go 15:1 just to make 1 point on the ES, even if it was for a few minutes. That’s stupid. I think you can go broke doing that. In a couple minutes you could be down a few points, when you were just looking for a nice, easy 1 point to take home and brag about. That few points turns into 8 points, and you start to get nervous, then you could be down 15 points in the next 15 minutes. Then, at 15:1, you start to question the validity of your system. And that’s what kills systems around here.
The other thing to remember is that unless they’ve been around for more than a couple years, you can basically write off any system with less of a history than that, even if it has increased 10 fold. If you drewdown 45% to do it, I think you’ve only really increased 5 fold, which is still good, but I wouldn’t put a dollar with it.
So to address your question, I guess if you think you’re overleveraged, take a smaller position. Offsetting it with hedges is only useful if you’re trying to keep a longer term position active. There are systems on here that have done that, but their Sharpes and APD suffer with the offsetting trades.
Think about why higher leveraged systems fail. While the vendor’s research might work for a few weeks, it goes through the first real-time drawdown and turns out much worse than originally anticipated. That first drawdown is enough to make people quit for good.
Maybe Jon was right that nearly all systems are discretionary, but I would hope the ones that are still around have a solid backtest that acts as a support when they start to committ real money to it.
Beau,
Thank you for taking the time to write out that detailed explanation. I am sure that many on this site besides myself will find it useful. After reading it, I believe I am seeing your original point a bit clearer which is that, calculations for margin interest are irrelevant because highly leveraged systems that will benefit the most by not having the fee, won’t be around very long anyways. Is that about right?
It also addresses another question I have had for some time regarding the Sharpe ratio…My system, QQQQubert, although only using 2:1 leverage with a drawdown of less than 14% and an APD of .4 has continually had a low Sharpe ranging from .75 to 1.2. I always thought it was due to the way I use options and that they don’t fall within a normalized dataset on which the Sharpe relies. But I then came to focus on the fact that gains above the average are treated as severely as losses below the average producing a statistical variance that artificially increases the risk calculation for the Sharpe ratio. I still believe these factors are negatively impacting my system’s Sharpe ratio. However, your point about position sizing made me wonder if this perhaps has the greatest influence because I trade based on fixed sizing adjusted minimally for market conditions.
I have started a new system called GoldenEye to hedge QQQQubert. Both trade with the exact same methodology, but are highly non-correlated (-.74). Eventually, I will blend the two into one system, but in the meantime, I will continue to watch the stats and see how they average out when combined. Based on what you have shared, my guess is that the Sharpe ratios on both systems will continue to be low.
Anyway, thanks again, Beau. I wish you continued success in your trading.
Sincerely,
Daryl
[LINKSYSTEM_26717594]
