Discussion of article "Extract profit down to the last pip" - page 33

New comment
 

Forum on trading, automated trading systems and testing trading strategies

Discussion of the article "Scratching the profit to the last pip"

fxsaber, 2020.03.15 23:15

Super Profitability.

During a storm, the market creates a huge number of small imbalances between fundamentally related symbols. Those are the logical ones to trade. In fact, they are the only ones to trade.

I.e. the task of an algo trader is to create a safe harbour in the form of a symbol, which is quiet during the panic around.

It looks like a second wave after March 2020 is coming.

 
fxsaber:  Looks like a second wave after March 2020 is coming.

A second wave of what?

fall?

growth?

pseudo-virus?

 
...:

second wave of what?

The storm.

 
fxsaber:

Storms.

there's no scratching out profits or waves.

It's all dead.

 
There was mention of Sergei Kovalev, who's already drained the account.

So he's already drained it.

 
It's some kind of a "do-it-all" fryer.
 
Can I ask you a couple of questions about a hypothetical TC? I am interested in 2 points: 1. What is the average time of holding a trade on a hypothetical profitable set? 2. And what is the average depth of history for analysing / making one decision (trade)?
 

Maxim Dmitrievsky:
Можно пару вопросов по поводу гипотетической ТС? Интересуют 2 момента:

1. What is the average time to hold a trade on a hypothetically profitable set?

Very much depends on the symbol. I haven't looked at statistics for a long time. The average is probably about an hour. Depends on volatility.

2- And what is the average depth of history for analysing / making one decision (trade)?

I never do something like this:

Fixed patterns, and even from the timeframe. Moreover, I do not analyse the signals, but look only as a result of the TS.

For example, there can be a lot of sell signals. But if there is already a SELL, I don't care how good these signals are. Otherwise, I have to consider the TS in the form of fractions - not my thing.

 

The grit that ruins everything. White swans in history and in reality.


White swan.

When optimising the TS, you can run into such situations.


There is a total profit, but it is obtained on a very short interval. On the screen I have shown in detail - it is less than an hour (minute timeframe).

It is clear that there is no systematicity here, despite the plus backtest. This is just a white swan, which flew in because of a crooked indicative quotation or for some other reason. Tuning the TS on white swans is fraught. That is why, as a rule, white swans are tried to be cut: either they are simply banned from trading, or the story of a white swan is replaced by a grey mouse. In general, everything is done so that graality does not distort the result and does not interfere with finding patterns. Exactly the same is done with black swans - mentioned in the article.


The reality of the white swan.

But one always wonders what happens if one encounters this bird in the real world. Especially when the technical infrastructure of both the broker and the algotrader is at a very high level: absence of negative slippages at limiters, adequate processing of redirects by the broker, TS based on ticks without skips, virtual trading in real time and other tricks that can help even with HFT-trading.


Rollover.

Looking for the white swan all over the market. Statistically, it arrives most often in the rollover. The only thing to do was to wait. Paradoxically, the rollover is characterised by abysmal execution and wild spreads. And I wanted to scalp the bird, which is somewhat inconsistent with the terms of scalping at first glance. Full of complaints about how terrible it is to scalp during this period. But a small lot can be sacrificed for a unique experience.


Real.

And he was ambushed. Now we are only talking about real trading.

Rollover scalping. Detailed steitment in the attached file.


Let's take a closer look at the price behaviour. Here are the ticks superimposed on the bars.

It's too small and nothing is clear. And why do we need bars, if we proceeded only from ticks anyway. So let's look at the ticks in detail.

Here we can already see that there were very short-term upward movements of Bid-price and downward movements of Ask-price. There are not many ticks.

This is what a white swan looks like.


Result.

If you want, you can see all the details in the interactive trade report. I will comment only on this record.

Positive slippages more than seven times covered the commission costs. The average lifetime of a position is 72 seconds. The shortest - 154 milliseconds (with a ping of 41 ms).


Just over a hundred trades. Virtual trading shows almost 250 trades. I.e. very strong discrepancy - virtual is much more profitable. This is another confirmation why it is better to bypass white swans in Tester.


Don't do it like this.

Files:
Report_Rollover.zip  132 kb
 
crap saber is pouring)))



and someone said that he is the best trader.

if the best trader is pouring like that, what can we say about others.
1...2627282930313233343536
New comment