Mincer #21 The saying goes perfectly to trading: "Many went to shear the sheep, but returned sheared." This week, Wednesday and Friday, suffered major losses trading after 15:30. This is a good reason to analyze all ES trading. Obviously, MES trading is demanding on the quality of transactions. Any strong losses result in a stop on net profit. Last week, volatility fell and the problem showed its obviousness: Grosprofit in green, Netprofit - strongly in red. Let us analyze how the quality of trade will change if we imagine that we trade only ES with a minimum commission (3.88). Let us compare reality with a hypothetical version in one table: Separately, select the last month and two months: The conclusions are simple: 1. The commission at MES is very voracious. The quality of the curves changes completely and there is a significant margin for dips. Therefore, it is better to start trading with mini- and at a minimum commission. 2. It is better to start not with $7000, but with ~$15000. Then the drawdowns will be workers (up to 7-9%) and not undermine the balance of the psyche. 3. Now volatility is falling and tough times are coming - when, after receiving a loss, the market does not provide new opportunities for entry. It will become much more difficult to earn money in comparison with March-April. Because of the commission, on the one hand, you have to leave for ES, but on the other hand, be prepared for a 20% drawdown. Or combine: a volatile market to trade MES, and a fading market - ES. Now is the weekend, so I will continue to trade 3 MES and think how to live. P.S. A bug was detected in NT7: non-existent transactions are registered on a real account with similar to real IDs:
FDAX Forwardtest #16-#21 Using the weekend, I tested the setup on FDAX, which I have been eyeing for a long time. The entry point is very simple - two-parameter. Of those that are visible through the eyes. An interesting nuance - the asymmetry of the market is clearly visible. A good plus is that take profit is a bit more than stop loss. 29 days 169 transactions: +95; -74; 56%+. It turns out 234 points in 1.5 months for 1 contract. $4,500 per month with a drawdown of $2,000. Calmar 3.1. With a drawdown of 10% of the deposit, the trading capital should be $20,000 for 1 contract. Estimated yield is 20% per month. But the liquidity of the system is very limited. Naturally, more extensive testing is needed. But now there is already one technical problem - the time in the market is 14 hours a day, which is difficult for manual execution.
Made a calculation of the matrix RR and the probability of the target. Three options: - Real probabilities in my mind; - optimistic; - Super-optimistic. And the three options are the same with the trailing stop. Conclusion: - It's scary to live)) - As expected, long goals do not justify themselves in mathematics. The longer the target, the more carefully the deal entry should be chosen. But the range of the target and the possibility of scenarios along this path greatly underestimate the probability. - Trailing stop significantly improves the picture. - The most important thing: additional entry between the point of initial entry and the stop is mathematically much more profitable. If you always enter your trades 1 point better than planned with the same take-profit point, this can VERY significantly change the picture. Here is the file if someone wants to play around with their probabilities:
We model the ideal process: - takeprofit varies from 0 to 20 points ES; - the probability varies linearly from 100% to 0% in increments of 5%; - stoploss is either fixed (full) or trailing {-1, -2, -3, -4} pt; - fees and slippage are not taken into account. Question: what values of takeprofit give the maximum and optimal (-10% of maximum) profit? Here are the charts for the four stoploss values: Here is the summary table: Obviously, with a linear probability field, regardless of stoploss, the optimal takeprofit is 3-4 points. Therefore, the main and only task is the marking of zones and sections as a probability field. Which greatly depends on the method and approach. The closer to reality your view of the market, the more adequately applied models, the higher the chances of finding areas with a probability of 65-80% at an amplitude of 3-4 points. With a stoploss of 3-4 points and the well-known RR=3 without using a trailing stop, you have no chance. And often do you see recoilless movements allowing such a trailing stop to slip? But since the real probability field is different and shifted to thick ends, then, possibly, somewhere on the RR above 6-8-10 there may be a positive zone. But! It is very likely that in doing so we will get out of the RTH borders with all the ensuing hole in risk control. And then a dumb question comes to mind - if you can’t model the probability field for your high stop zone by -3-4 points, and at the same time try to simulate the territory by 10-20-30 points for your high take zone, then what your brain is trading and what for?
Mincer #23 Last week I bargained normally. But today it has severely failed. As a result, he rolled down to the starting deposit amount. Three months of trading empty. I'm at zero. Of the acquisitions, only a license for a reduced commission. Today I stopped trusting myself and introduced restrictions on risk management by the broker.
Mincer #23 I take several setups to trade from Vintage. I'll see what happens. In history, it looks like this: Low volatility: Average volatility: High volatility: