The future of automated trading: round two - page 5
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
There is one "but". We cannot know the future movements of the quotes. That is why the conclusion is wrong. One does not follow from the other. Precisely because we don't know future quotes, much less to the 6th digit. )))
You don't need to know. And no one does. We are talking about probabilities in any case. The analogy with the weather is bad because there is no mutual influence, but let it be. Weather forecasts don't guarantee anything
or promise nothing, as no one knows what tomorrow's weather will be like. But in general, short-term forecasts work.
If the road itself is discrete, why not drive on a discrete road also discretely?
And driving backwards, looking not forwards but backwards and guessing from the road we see behind and have already passed, what and where the turns will be is not suicide? )))
Well then for you I suggest the other extreme. This is how hypotheses are tested to see if they work in boundary conditions. You have to work with candlesticks, annual, once a year you open a chart, make a deal and rest, that is how you see the future of ATC?
I don't know how to say it, it's so obvious, understandable and logical for me ... Maybe someone will at least read and understand, and will not optimize their system for a month ...
Let me explain well on the third side. There in EI theory the mathematicians have shown that to achieve the maximum (minimum) of the target function. This could be, say, the maximum growth of your deposit over a year (a year is the time interval of control). You need to know the FUTURE.
Knowing the future movement of quotations within 6 digits you personally can build a profitable TS? Now out of all the many profitable TS which one is better? which one is controlled once per hour, or controlled constantly? which of these two systems will beat the other? which robot is better? more profitable ...
You want to drive in a car that is controlled once an hour, go ahead, I can't stop you. But remember, there is a better one, a priori better, it's the one who drives the car all the time ...
To continue the analogy, let's consider driving a ZIL-130.
Power steering implies reasonable backlash, and many drivers try not to miss the road and use jerking of a wheel, those back and forth motion.
One can not drive fast with such car not because it has not enough power (you can overdrive with a better gear) and because at high speed critical requirements for steering system will be increased. There is only one way out - and that is to reduce the speed. In our case it means increase of time between control signal and new forecast with latest data (response). Moreover, the lower speed (the greater the discreteness), the more likely you are to reach your destination.
Answer the question on what is more likely to reach your destination :
in a Lamborghini Diablo? Or
a paving roller?
A paving roller, though slower, is more likely to get there even if driven by a drunken watchman instead of the driver :o)
But in real trade we choose laborgini as it takes time to drive it to diablo and back, so it's possible to chop more cabbage.
If the market is in the middle, then wherever you open, you get profit (if you have enough patience to wait) this is a rink.
And this strategy is opposed to scalping (lamborghini diablo), all the rest is a transitional model.
A 10% rink on 90% diablo, or vice versa, are just variations of the same process.
The higher the discreteness the lower the requirements for the forecasting horizon, but also the higher the requirements for the execution system.
A sensible middle ground always wins.
Continuing with the analogy, let's look at the steering of a ZIL-130.
...
...knowing the future movement of quotations to the sixth digit, can you build a profitable TS? I think so. Now out of all the many profitable TS which is better? the one which is controlled once an hour, or constantly controlled?
Of course everyone would agree that if there is a magic system knowing the future and working with zero spreads, it is better to trade this system on tick data to make money on all small price movements. I guess the discussion was started when some participants claimed that the small traders will die, because with their big spreads, low-quality quotes and slow computers they cannot compete with Goldman Sachs. In this case, for some reason autotrading was understood as pipsing. In my comments I was opposed to such pessimism saying that it is possible to create a successful trading system operating on hourly or daily quotes. Now I see that the discussion has moved to another direction: trading on what timeframe is more profitable. It all depends on spreads, liquidity, frequency of successful trades (predictability of the market at this timeframe), the speed of code, and other things. I will not argue here, we are all different. But I will not say that the autotrading of small traders is going to end either.
Prival's statement gave me an idea. There are a very large number of attempts on this forum to apply various mathematical tools to the market.
And as a rule, the vast majority of these attempts are unsuccessful. And all these failures can be explained by one feature of the market, its nonstationarity.
The attempt to apply mathematical models for stationary processes to non-stationary ones. So here the question to Prival, that universal control model for which processes?
Imagine a trader knowing the future price movement to the sixth digit, possessing a billion-dollar deposit, betting for tens of thousands of lots, what will happen to
the market. The only thing I can say is that it will not end well.
You don't have to go from one extreme to the other )))) Why the annuals? The truth is somewhere in the middle ....)))
Prival : I don't know how to bring home to me what's obvious, understandable and logical ... Maybe at least someone will read and understand and won't optimize their system for a period of time ...