Econometrics: one step ahead forecast - page 101
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
You should read more carefully what you are being asked yourself. You didn't show the whole ACF, just the first 500 samples, even though the sample was larger. That's what I was asking about.
Yeah, I read you carefully, you took a sample of 500 counts.
Show me all of it, all 5,000 samples. I'm just wondering what it's gonna look like.
Look, I'll get to the lab and if I remember, I'll show you. You can look at the ACF for any number of samples.
My research indicates that about 90% of it is an oscillating link model, if you do the right processing, although you don't have to...
what the fuck is an oscillating link... what are you, an econometrician too? Tell me you're normal oriented, one faa on this forum is enough for me...
Don't read this thread and you'll be all right in the head.
But I asked you not to answer, you are violating the hard-won consensus. You are an econometrician to the bone, but where did I write about the head? There are also nerves, for example, and other realities that are given to us by sensation. And how can you be so sure that your head is all right. Why the arrogance? You run around the forum with 40 tests made on a model that is not suitable for predicting quotes, you write nonsense and talk about normalcy. For you the best doctor is the market. Doctor's appointment 24 hours a day, 5 days a week: from 01:00 (Moscow time) Monday to 01:00 (Moscow time) Saturday.
PS: But I hasten to note, nevertheless some kind of humanism in you remained, you recommend not to look at this topic. Apparently, in the literal sense, I will take advantage of your offer. It's very tempting.
faa - based on your spreadsheet I have made my own -
date/forecast price/forecast price/real price/forecast error/profit
The average forecast error is rather significant - about 119 pips per forecast. But it turned out that if we trade according to these though inaccurate forecasts we obtain profit in general. I calculated profit according to the forecast - in the forecast point I opened a position with TP equal to the price of the forecast, without SL, the losing position was closed at the next opening price. It's amazing that analyzing only 4 bars you get such MO, (unless I messed up).
faa - based on your spreadsheet I made my own -
date / price at time of forecast / forecast price / real price / forecast error / profit
Average forecast error turned out to be rather significant - about 119 pips per forecast. But it turned out that if I trade by these though inaccurate forecasts I get profit in general. I calculated profit according to the forecast - in the forecast point I opened a position with TP equal to the price of the forecast, without SL, the losing position was closed at the next opening price. It's amazing that analyzing only 4 bars you get such MO, (unless I messed up).
I have this last column. I think the available forecast is not that bad. Profit in pips is questionable, as it is too reflective of a particular kotir area. More interesting is the profit factor in observations = number of profitable trades/number of losing trades.
But this is not the point. At this stage I'm not interested in profit - I'm interested in stability of the model, not even in stability, but in model parameters that will help me to judge about its stability in the future. As they say, feel the difference.
And so I can make profitable models, but what guarantees future profits? Only the stability of the model in an unsteady market
faa1947: Более интересен прфит фактор в наблюдениях = число приб сделок/число убыточных сделок.
This is not how the Profit Factor is calculated. It is a ratio of "number of profitable trades*size of average profitable trade/(number of losing trades*size of average losing trade)".
That's not how the profit factor counts. It is the ratio of "everything earned" to "everything lost".
OK, I agree. The result will be the same, but why perform three calculation operations on four independent data, when the same result is given by one operation of the relation between two data? You mathematicians always complicate things :)