Previously I have discussed the relationship of the kelly criterion and correlation. Previously I have also shown that MORE bets at a lower correlation (ideally zero) is often much better.
The problem is FEES. Previously I had not quantified the difference that fees make in your portfolio. Now I have made the adjustments to the spreadsheet so that you can. You would be surprised that even with a LOT more capital, you still will be limited by the fees.
10 bets at a time, no problem, right?
One of the best traders I know who in a very short time frame can produce an average of 27% returns 26% of the time, 7% 20.61% of the time, and a "scratch" (0%) 10% of the time and most of the time capping losses between 5 and 10% with the occasional 15% loss. I figured, someone just starting out could easily play 10 bets at a time, right?
WRONG! With $5 per trade costs, it takes a bankroll of about 100,000 until 11 bets is better than 10 bets and only very very marginally so. AND THAT is at a correlation of zero. Put 70% correlation on it and you need 150,000 before you should even consider 10 bets.
I was shocked, especially considering I have been marginally profitable this year trading more positions than that. Granted, I used options, but the position sizes were even smaller so the fee ate me up even more.
Especially considering the kelly often favors underbetting, and the only way to do that without causing fees to eat up your bankroll to a greater extent, is FEWER bets with more capital at risk. I now wonder how much of the success was just varience, another word for basically saying "luck". I will reform my strategy immediately. no more small trades, and no more portfolios with so many positions... Until I build my bankroll, it's time to apply focus.
It's a bit scary because it is really leveraged to the results of a very small number of trades over the course of a year. That doesn't matter, the fact is the math suggests that it can handle the varience better because of the fees impacting it less.
I have dramatically over estimated the true "cost" of fees and the starategy it takes, and instead opted for a lower volatility, marginally profitable to slightly negative strategy.
With this style of trading and 2 single bet at a time, how much does one need to get started?
With a correlation of 70%, until you get $2500, you should not bet more than 1 at a time. At $2000, 1 bet is better. If correlation is 50, you need $1750 to run 2 bets at once. If you have 5k you probably only want maybe 3 bets at a time. I would still be one less at a time than suggested because that's the only way to apply less risk without the fees eating you up and reduce the insane volatility of the strategy. If you dip below that, you have to start saving up before you can resume again.
If the best swing trader I have seen can't do it with those numbers, neither can you.
But You can do options, right?
Not so fast. Options are more leveraged, so the "no fee" optimal strategy is to risk less. If you risk less and fees are the same (plus an individual fee per each contract), you need more capital. For example, 2% of a $5000 bankroll is only $100. $10 fees ($5 buying and $5 selling) means 10% per trade. That's a HUGE barrier to overcome, even with substantial leverage. Additionally, there is another disadvantage you have to overcome. The difference between the bid and ask. The spread is usally another $5 per contract. However, if you can find the ideal trade, you may be able to make it work. Please note that buying a ton of excess time such as a year can really be helpful because it limits the risk. Thus, you can put more capital at risk (than out of the money options), thus fees can impact you less. Unfortunately the yearly contracts typically have the widest bid-ask spread of all of them. So you theoretically could do options IF it was a very, very good setup in a very good stock that has very liquid options and you can get a fair price... AND the expectations are still as high as the stock trading strategy. Since many of those stock trades were small low priced volatile stocks, you are probably going to get a few possible trades a year that work with this strategy.
STOP TRADING, start investing
The advantage of investing is much more significant for a small trader. An investment may only produce maybe something such as a 20% gain over a year depending on your skill level, but when you consider you avoid the compound costs of constantly going in and out of a trade, it's likely worth it. A DRIP (Dividend Re-Investment Program) Helps you even more if you are lower funded like most retail traders because you get to automatically reinvest the dividend without FEES and effectively dollar cost average in without paying fees. If you are going to have so many positions a once, you are better off sticking it all in the SPY. Better yet, take $500 in the SPY, $500 in the GLD, $500 in the TLT and $500 in cash to get started, and then build upon it from there, particularly if you have a "commission free trades".
Bottom picking for small fees... sometimes.
Another strategy is "picking bottoms" You basically look for an oversold stock in an oversold sector (using RSI below 40 (or 30), with fundamentals worth buying and holding. You buy on either a candlestick popping up and closing above the high of the previous day candlestick, or else you look for other signals. Regardless, a close below new lows and it's easy to exit and not convince yourself there is a reason for holding on. You will have several failed trades, and some that pop 10% before declining again, but you are looking to just hit one great trade in which you hit the bottom and can let it ride and pay for all your losses and then keep going.
Conclusion:
Trading is hard enough as it is. Add fees and it is even harder. The "little guy" still can make it work if from the start he knows what he is doing, he trades extremely well, he limits himself to one or two trades at a time, and he manages his trade with discipline and grows his account over time (with discipline). You should not approach it as "easy" like all the infomercials suggest, but incredibly difficult especially with a huge disadvantage of a small account. As a result, ONLY strict discipline and a lot of knowledge and experience will give you positive results.
I believe there are some solutions that I will be able to find to counteract some of the drawbacks, but it is not going to be easy to build your account.
No comments:
Post a Comment