I'm interested to hear if global traders and investors normalise the price of foreign assets (equity, future, etc.) from its quoted currency to their own base currency. For example, suppose I'm based in the UK and my funds are in GBP and I want to compare a Canadian energy company quoted in CAD to an Australian energy company quoted in AUD. I normalise the prices from their quoted currency to my base currency which is GBP, i.e. I convert the Canadian company CAD prices to GBP and the Australian company AUD prices to GBP. Then I compare the companies on an apples to apples comparison basis and use the converted prices in my strategies. Does anyone use this type of technique, if so why, if you don't then why not?
It depends on what type of thing you are doing. When position sizing you absolutely need to convert currencies. Otherwise your CAD or AUD positions would be undersized compared to your GBP. However in a lot of cases it doesn't make sense or makes no difference. A trivial example is if you're using dividend yield. Dividing dividends in CAD by CAD price and the currency cancels. On the other hand if you have a company that pays USD dividends on a GBP price (eg my ex employer EMG.LN) then you need to a conversion. An example of where it seems to make no difference is if you're trend following; it doesn't matter if you measure the price in CAD and then do all your moving averages etc etc, or if you convert the CAD price into GBP first and then do your charting analysis. FWIW in this situation I never convert the price. For a more specific answer it would help if you described what you are doing in a bit more detail. GAT
@globalarbtrader I completely agree with the position sizing, you'd want everything in a common unit of risk. Lets say I'm momentum trading baskets of global equities long/short. I'd like to measure momentum of each equity in a common unit so I can compare. When I mean momentum lets assume I'm talking % return over a period. The question is what difference, if any, would it make to convert all prices to a common base currency, then measure the momentum, as opposed to measuring momentum of each equity in its quoted currency price? I assume that this gives an additional dimension to trading and investment in that equities in countries with a currency that is appreciating in comparison to your base currency would get extra attention. Surely this is a good thing?
It depends. I guess it depends on what you think causes momentum and who the marginal investor is. If the marginal investor is foreign capital then perhaps its true that you'll be capturing momentum in both the currency and the stock at the same time. If the marginal investor is domestic capital this seems less plausible. Another argument is that people tend to look at charts in the same currency as the stock is issued in. So even foreign investors look at charts when making investing decisions that don't account for currency appreciation. Like I said when I tested this I found no difference between these two measures. So feel free to convert things into base currency first, but you probably won't make any extra money. For me personally it makes more sense to trend follow most things in their local currency, and then trend follow currencies separately. GAT
@globalarbtrader Thank you so much for the feedback. From your perspective I guess it makes complete sense to follow assets in their quoted currency, since as you state you trend follow currencies separately. From my perspective, where I long/short a portfolio of global equities, it makes more sense to convert equity prices to a common unit and then compare. Additionally, it makes no sense (at least to me anyway) for an investor with funds in a local currency to buy a foreign asset without taking into account it's 'real' price, i.e. the price in your local currency. Anyone, please feel free to propose a rational argument against this. In the meantime, I'm going to run some backtests and see what difference (if any) this makes. On a side note, it also brings us the question as to why US investors are so focused on US companies and ignore foreign opportunities (yes, I'm aware of global equity correlations, etc)?
If I pull up a chart for AAPL it's priced in dollars. It's a real pain in the neck to get a chart for AAPL in GBP. People who look at charts, and basically make momentum work, don't look at charts repriced in foreign currency. home bias cognitive effect tax and administrative hassle of owning stuff in different countries pain of having to translate currency to buy, and back again availability of ETFs for cheap and straightforward investment overseas difficult to get reliable data especially for emerging markets or if in foreign language US market is large and relatively diversified so costs of not going overseas are relatively low GAT
Yes - costs and data are real issues. Especially, if low cost alternatives like ETFs capture the majority of foreign factors. Some of these less accessible foreign markets are less efficient and thus present greater opportunities IMHO.
Obviously yes for sizing. No for monitoring or analysis. It is useful to see if returns came from price appreciation or from fx movements, see if I am happy with local price, current FX rate and expected correlation. For example, looking at Russian MICEX index gain of 45+% in 2016 seems rather pointless without breaking it down into ruble appreciation of over 20% (one-off event - oil price recovery, inflation stabilized) and 25% from performance in local currency (inflation down -> lower bond yields -> higher bond prices + econ bottomed out, imo). Going forward, I'd expect that further ruble strength could hurt exporters and econ recovery (non-oil exports like steel), so upside is more limited. Looking at fundamentals in your home currency seems intuitively challenging for me. Sberbank (AAPL of the Russian stock market) profits took a huge hit in USD as ruble fell from 35 to 65 in 2014-2015. It's difficult to decipher how bad the deterioration in profitability is by looking at USD numbers of a company that operates domestically with a depreciation currency. It's also a nightmare for modeling stocks. Practically it is difficult to track instruments in home currency when order books, trading platforms, analysts, news articles and etc use the native currency unless ur a prodigy capable of instant convertions in your head.
Yes, I do something very similar. I decompose an asset into the asset in terms of a standardized numeraire and its base currency in terms of that same numeraire, then analyze and forecast separately. Then re-aggregate the individual forecasts for trading. I use the same numeraire for account valuation.
I completely agree, looking at fundamentals in your base currency presents challenges. For, example the depreciation of GBP in 2016 due to Brexit has lifted FTSE100 multinational companies since the majority of their earning come from non-GBP sources. However, they report their fundamentals in GBP. Converting to USD for a US based investor presents challenges. For example, what exchange rate do you use? Date of the reported earnings? Beginning of the year? Nightmare. Dividends are more straightforward since they have a specific date. For this type of asset price analysis and forecasting it is relatively straightforward if you utilise your own programmed software - which is a must. Interesting, so rather than use the base currency per say, you use an intermediate representation normalised by your "standardized numeraire". I'd be interested to hear how you do that and why you don't just use raw currency conversion?