Thoughts on some of the absurdity of multi-currency analysis. - page 7

New comment
 
getch писал(а) >>

The fact that EURGBP == EURUSD / GBPUSD is ALWAYS correct with a spread adjustment is 100%.

Above gave links to EAs that show this perfectly in practice.

The man wasn't writing about the correlation between synthetic and currency pair.

He was referring to currency pairs. Everything has been clear about the synthetics a long time ago. And the difference there is not even in the spread as shown by the aforementioned Expert Advisors. All the arbitrage opportunities in MT4 arise only because of the curvature of quotes of brokerage companies, in the reality the "spread" is even smaller.

There are three components to synthetics. So if we accept that there is a stable correlation between pair 1 and pair 2, then there is always a 3rd pair that can behave as it pleases.

The hypothesis that there is a stable correlation between 2 pairs automatically means that you can accurately predict where the rate of the third pair is heading now. That is why there is no sense to play on correlations, you can just chop some money on the third pair.

 
Zhunko >>:


В общем смысле неправильно. Так, как вычисляю, ещё никто не догадался. Не считая моих компаньонов.

У меня динамические весовые коэффициенты.

EUR/GBP should not be confused with my EUR/my GBP index. They are completely different concepts.

 
vasya_vasya >>:

Человек писал не о корреляции между синтетикой и валютной парой.

Речь шла о валютных парах.

Then you've got it wrong.

 
marketeer >>:
Когда мы рассматриваем только 2 пары, то да - трудно сказать, какая из них перетянет другую. Но когда у нас с одной стороны портфель пар, а с другой - некая пара, ведущая себя "неправильно", то я б предположил, что вероятность вытягивания её в сторону портфеля выше.

Your intuitive assumption will be correct when the market has the specific properties attributed to it by portfolio theories. However, this is not the case. The market is not a spherical horse in a vacuum, so even the most advanced portfolio and arbitrage theories, written by Nobel laureates in economics and verified by trading gurus, almost immediately stumble upon failure.


Getch wrote(a) >> A portfolio of currency pairs does not diversify risk, or rather it does not reduce it. - Quite right.

 
getch >>:

Не надо путать EUR/GBP с мой индекс EUR/мой индекс GBP. Это совершенно разные понятия.


Well, of course...! And the indices are calculated from what? A lot of currency pairs.

So multicurrency analysis is not hopeless?

 
Vita >>:

getch писал(а) >> портфель из валютных пар не диверсифицирует риски, точнее, не уменьшает. - Совершенно верно.

Example:

There are two profitable (on history) strategies. One on GBPJPY, the other on EURUSD.

Will the risks be reduced if instead of one strategy, we run both of them at once?

Of course, it is assumed that the deposit load is the same in both cases.

Answer:

Intuition suggests that the risk will be less if you run two strategies instead of one.

But the points made in this thread above say that this is not the case.

P.S. Started using the text format, as it becomes much more clear and systematic.

 
Vita писал(а) >>

Getch wrote >> A portfolio of currency pairs does not diversify risk, or rather it does not reduce it. - Quite right.

In my opinion it does not. Diversification is effective in any case where we are not using the same instrument to create a portfolio. This thing works almost always.

If you have any arguments against its use - post them, it will be interesting to read.

 
Zhunko >>:


Ну, конечно!... А индексы из чего расчитываются? Из множества валютных пар.

Стало быть, мультивалютный анализ не безнадёжен?

You construct your currency indices from an analysis of multiple currency pairs. Then you compare their ratios with the corresponding exchange rates. And of course you see the differences.

This approach does not say anything about the hopelessness of multicurrency analysis and vice versa.

 
getch писал(а) >>

Example:

There are two profitable (on history) strategies. One on GBPJPY, the other on EURUSD.

Will the risks be reduced if instead of one strategy, we run both of them at once?

Of course, it is assumed that the deposit load is the same in both cases.

Answer:

Intuition suggests that the risk will be less if you run two strategies instead of one.

But the points made in this thread above say that this is not the case.

P.S. Started using the text format as it becomes much more clear and systematic.

The point is that diversification into two different strategies is a foggy thing until you see a graph of change in equity of these 2 strategies.

The fact that the strategies operate on 2 different pairs indicates that the entries may occur at different times, respectively, the loss of the 1st strategy will not always correspond to the loss of the 2nd one, which is the effect of diversification. The total drawdown would decrease in this case. But of course it is not a fact.

As for strategies, they can use the same effect synchronously making deals simultaneously and in this case no effect might appear.

 
getch писал(а) >>

ECN and crosses:

On an ECN, the crosses are initially virtual. That is, their liquidity is calculated from the liquidity of the majors. Naturally, ECN clients may influence (towards narrowing) the spread with their cross trading orders. Such orders are overlapped by the same real opposite orders, or virtual ones (formed of the majors).

CADJPY:

It would seem that liquidity in such a specific cross as CADJPY is quite poor. However, it is not so. When an interbank marketplace decides to introduce an exotic cross, it fills it with virtual liquidity through the majors. On ECN, any cross is initially made via majors. Further filling of liquidity of the cross happens due to trade requests on the cross by participants of ECN.

ECN and STP:

In general, the ECN scheme is quite specific. Not all liquidity providers guarantee execution at their prices. I.e. they cannot be a priori participants of ECN. However, they participate. It is done (by the ECN-provider) through an artificial worsening of prices (spread widening), so that the ECN-provider can guarantee the execution. Moreover, quotes in ECN are filtered to show the appearance of perfect execution. For example, a liquidity provider threw in ECN some orders that should have overlapped with yours. But the ECN could not (lots of reasons) overlap the counter orders. You will not see that there were bank's orders. I.e. in ECN the execution takes place first, and in case of success the prices are updated. Otherwise - not.

In this sense STP is much more transparent, however, it is trading on an indicative without any guarantee of execution.

Liquidity providers may calculate their bids and offers however they want, even through majors - it is their business and there is no regulation about it. On ECN and exchange, everyone can create liquidity by placing orders within the current spread (narrowing it). Therefore spread on crosses does not depend directly on the major and may be much less than the sum of spreads on corresponding majors.
1234567891011121314...40
New comment
Reason: