Confused by the inflation news - effect on Forex?

[Trader200K]

Yes, I've been coding/testing indicators/systems for more years than I like to admit. I enjoy it. While I make a little money trading, I'm still a better coder than trader LOL. Max stated my conundrum perfectly a few posts back ... "trading and logical thinking/reasoning don't mix".

I am going to be the first to sign up for Maxinger's Trading Course if he ever offers one. Part of me viscerally gets his point, but I logically don't understand why . Clearly being bred as an engineer/quant trying to trade is a major mindset handicap.

If you have any ideas for custom indicators, pm me and let's chat. I've been rolling around how to best meld multi-timeframe data into a indicator/system to improve consistency of entry/exits, etc.

I'm always up to learn/try something new that might improve my trades as well.

 

[hanneswas]

Using price is not a good analysis. Figure out the driver of the pair and what the actual risk is, and model the impact to the exchange rate. FX pairs do exhibit momentum but if you don’t conduct an analysis you’ll be the dumb money providing liquidity for the informed.

Would you mind to help me with an example of such analysis you referring to? Some article maybe?
Thank you!

 

[longandshort]

Would you mind to help me with an example of such analysis you referring to? Some article maybe?
Thank you!

What's your background?

Here's a good primer for why reading the news is too late:
Who Makes Money in the Stock Market (The Only Game in Town) - YouTube

I recommend that you read up on principal component analysis and pick up R. Then you should brush up on the International Fisher's Effect, the law of one price in economics, and the carry trade. Once you understand these, you can begin to model.

PCE --> tells you what factors are driving a pair

For example:
USD/EUR (1-Month %) = factor_a*weight + factor_b*weight +...+ factor_n*weight + error

You can express your view on the factors you have an edge on and weights you think will occur. Factor weights are usually time-varying.

Then you can compare that to 1) FX forecasts and 2) current exchange rates. You can also model the impact of rates (assuming it is factor_a) to you can test various scenarios. This is what professional traders do.

 

[hanneswas]

What's your background?

Here's a good primer for why reading the news is too late:
Who Makes Money in the Stock Market (The Only Game in Town) - YouTube

I recommend that you read up on principal component analysis and pick up R. Then you should brush up on the International Fisher's Effect, the law of one price in economics, and the carry trade. Once you understand these, you can begin to model.

PCE --> tells you what factors are driving a pair

For example:
USD/EUR (1-Month %) = factor_a*weight + factor_b*weight +...+ factor_n*weight + error

You can express your view on the factors you have an edge on and weights you think will occur. Factor weights are usually time-varying.

Then you can compare that to 1) FX forecasts and 2) current exchange rates. You can also model the impact of rates (assuming it is factor_a) to you can test various scenarios. This is what professional traders do.

Thank you for the explanations!
Indeed I am familiar with the statistical analysis.
Sounds like "trading the residuum" between the prediction and the actual rates?

What are the factors you could recommend? Like returns from other currencies?

 

[longandshort]

Thank you for the explanations!
Indeed I am familiar with the statistical analysis.
Sounds like "trading the residuum" between the prediction and the actual rates?

What are the factors you could recommend? Like returns from other currencies?

Start with macro factors like economic performance differentials, interest rate differentials, inflation, etc. you can then add other factors such as momentum and sentiment.

 

[maxinger]

Yes, I've been coding/testing indicators/systems for more years than I like to admit. I enjoy it. While I make a little money trading, I'm still a better coder than trader LOL. Max stated my conundrum perfectly a few posts back ... "trading and logical thinking/reasoning don't mix".

I am going to be the first to sign up for Maxinger's Trading Course if he ever offers one. Part of me viscerally gets his point, but I logically don't understand why . Clearly being bred as an engineer/quant trying to trade is a major mindset handicap.

If you have any ideas for custom indicators, pm me and let's chat. I've been rolling around how to best meld multi-timeframe data into a indicator/system to improve consistency of entry/exits, etc.

I'm always up to learn/try something new that might improve my trades as well.

no maxinger course.

I started trading with hundreds of indicators.
after exploring for many many years, I found that the best is no indicator

 

[longandshort]

Start with macro factors like economic performance differentials, interest rate differentials, inflation, etc. you can then add other factors such as momentum and sentiment.

Follow up on this:

Economic differentials -- use PMI spread as a starting point (note, you may need to tweak/make adjustments based upon the type of economies you are dealing with). E.g. PMI of country A - PMI of country B.
- More advanced technique would be observing z-score and relative distance.

Interest rate differentials -- get yield curve data (deposit rate, 2 yr, 5yr, 10yr) and calculate the spread (country a deposit rate - country b deposit rate)

Inflation -- CPI spread

Momentum -- G4 Ahmed, Naka and Wuthisatian.pdf (academyfinancial.org)

 

[adonisperry]

@longandshort The forex market is definitely driven by these economic factors but I think Inflation is very confusing and its effect on currencies is hardly straightforward. You never know if it will place a downward pressure on a currency or if central banks will increase interest rates to curb rising inflation levels.

 

[longandshort]

@longandshort The forex market is definitely driven by these economic factors but I think Inflation is very confusing and its effect on currencies is hardly straightforward. You never know if it will place a downward pressure on a currency or if central banks will increase interest rates to curb rising inflation levels.

Agreed -- which is why we know it is a factor but we don't know the importance (weight) of the factor, nor have we done a qualitative analysis of different scenarios (central bank rate decisions). Hence the use of statistical models -- rolling regressions, correlations, and PCE to help us understand the time-varying nature of relationships. As a trader, you need to understand the historical significance of factors, where they are today, and then use a scenarios approach to figure out what may happen in the future.

For example, say a pair you're tracking is surging, despite recent high inflation data. If you look at the rates futures market you might see that the 2YR bond future is pricing in 4 rate hikes, which is 1 more than expected. Does this move have room to continue rallying? Well, if the rate contribution to the FX pair historically has been 40% during rate hikes, and is currently trending upwards from 20% to 30%, you can build a scenario where the rate contribution hits 40% to model the impact to price.

Now you have a trade idea -- the FX rate has room to run! You may have done an analysis on what drives correlations around rate hikes and found that most of it has to do with the view of international fixed income managers, who have all recently supported the move in rates. So as long as the underlying data (typically economic, employment, etc.) they are considering remains in trend, they will continue to be an incremental add in flow, which will strengthen the correlation between rates and the FX pair. Now you know what you need to watch out for, and you may understand the range of FX movement through the period better.

Then ofc you need to think qualitatively about it and make sure you aren't missing anything. But this is the general idea, and how global macro and fx traders conduct analysis.

 

[Trader200K]

no maxinger course.

I started trading with hundreds of indicators.
after exploring for many many years, I found that the best is no indicator

With experience of 100s of indicators, did you find them all unprofitable?