Econometrics: one step ahead forecast - page 6
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
I'm talking about the TC itself. After all, the bad equity is buried there. My main complaint about TA is that you can't tell what's bad inside the TS if it's bad.
No, it's not the TA that analyses equity. How to analyse - either use expert methods like various indicators like FS, Sharpe, Sortino, etc. Or statistical - the distribution of equity increments. The same HP tests that you applied. You can use chi-square, but how many of them are there?
No, it is not the TA that analyses equity. How to analyse it is either to use expert methods such as various indicators such as FV, Sharpe, Sortino, etc. Or statistical - the distribution of equity increments. The same HP tests that you applied. You can use chi-square, but how many of them are there?
There is some confusion.
The trading system generates equity. If we are not happy with equity, it is not clear what to look for in a trading system that has generated equity.
TA is a different story.
There is some confusion.
The trading system generates equity. If we are not satisfied with equity, it is not clear what to look for in a trading system that has generated equity.
TA is a different song.
your tests won't tell you what's wrong with TC either.
There are several aspects to TS building. There is a stage of evaluating the TS. The residual HP test that you perform is the stage of evaluating the TS. You just put it in before you calculated other metrics like profitability. You also used TA in constructing the TS. That Hendricks, or whatever it is :)
Your tests won't tell you what's wrong with the TS either.
I have already started to show in the topic - the probability of zero regression coefficients. My articles show more and will continue to do so in this thread.
I keep talking about the technology of building a TS where each step is evaluated, there are evaluation criteria and tools to calculate those criteria. TA cannot be used in this way. Building TS in TA is an experience + intuition, if something is wrong, you can only find out by equity, and then what to do - you don't know.
I have already started to show in the topic - the probability of zero regression coefficients. My articles show more and I will continue to show it in this topic.
I keep talking about the technology of TC, where each step is evaluated, there are evaluation criteria and tools to calculate these criteria. TA cannot be used in this way. Building TS in TA is an experience + intuition, if something is wrong, you can only find out from equity, and then what to do - you don't know.
So who should we change Hodrick and Prescott for - what does econometrics say? :)
In econometrics, it's the other way round. The model does not work. It is clear why it does not work. But what to do with this "understandable" - I do not know).
It hasn't happened so far. If the model stops working, it means that some test on a new sample fails for some reason and you have to start from scratch. It is the same with TA but unlike TA we can always diagnose the model before it enters the market.
You must agree that it is a principal advantage.
I see. Would not it be more logical to build a profitable TS and check the distribution of equity increments (not the balance) for normality as a selection criterion? Otherwise the cart is ahead of the horse :)
P.S. Perfect Equity is a random stray with upward drift. The distribution of increments is normal and the smaller the dispersion, the more grail))))
P.S2 And it is important that equity drawdowns do not have thick tails (one of the characteristics of NR), but let them be upwards.) For example, trend-following systems will have such spreads. The Sharp index estimates the equity quality on the basis of dispersion, but Sortino already takes into account only the dispersion of equity movements downwards
This has not been the case so far. If the model stops working, it means that some test on a new sample fails for some reason and you have to start building it again. It is the same with TA, but unlike TA it is always possible to diagnose the model before it enters the market.
You must agree that it is a principal advantage.
well, who to swap Hodrick and Prescott for - what does the econometrics say? :)
Change staff or beds?
Once again the general technique: take the initial quotient and see what can be formalised in it. Always trend and bias. Formalized, not necessarily with Hodrick-Prescott. Other methods are much more common. We look at the residual. Is there a trend? If not, try to model the residual, etc.