The stock market of the nation of Turkey, which has been a very poor performer this year:
(Chart courtesy of TradingView. Blue is Turkey returns; red is a broad portfolio of nations, excluding the US.)
Cap weighting would have had a low weighting in Turkey because it tends to favor nations with markets that have growth over long periods of the past. The cap weight of a market is basically the sum of the historical increases in market valuation over decreases in market valuation, so the US, UK, Japan, France, Canada, and Germany are in the top ten of standard cap weight indices such as ACWIX's index. Turkey is a newly developing country, emerging market official but just a notch above frontier. As of this writing, the market cap of its entire index is roughly $35 billion dollars. Exxon Mobile is ranked roughly tenth largest company in the world and the market cap of this company is ten times the market cap of the entire index of the Turkish national market. So cap weighting will tend to underweight Turkey severely. So pretty much any methodology other than cap weighting is going to have a strong tendency to overweight Turkey compared to cap weighting.
So then the next natural question after that is: Why have a weighting higher than cap weighting? In other words, what was the case for Turkey at the beginning of the period? Was there any good reason to invest there?
Mainly the case for Turkey’s markets was based on the fact that Turkey’s valuations tend to be low compared to other nations. It had been treated by investors as high risk and that means its valuations tend to be low in general compared to developed countries. This tends to be true of emerging market nations in general. EM is perennially 'on sale' and it also tends to overperform in the long run. Earnings, and dividends, come cheaper in EM than they do in DM.
If we look at Turkey at the end of June, we see that policy factors have been trending towards pro-growth economic policies (civil liberties are a different matter) and attractive valuations historically and correlate with higher than average total returns. But as we've seen, that has not happened. Turkey has been both a bad performer YTD and since the data at the end of June.
What happened? Part of what happened is that these models are not matters of certainty; they try to increase the probability of a good return. Certainty is unavailable anywhere in the world of markets.
In addition, there are factors such as political instability which don't usually matter very much, and then suddenly matter more than anything else. Data-driven portfolio construction tends to look at the things which usually drive returns, rather than exceptional events. The types of countries which tend to have political risk events also tend to be the countries which give the highest returns in the good years. If we throw out countries with political risk, we're throwing out occasional likely future melt-downs, but we're also throwing out more common future melt-ups.
Turkey was attractively valued at the beginning of the year. Why? Because Turkey had a coup attempt last year (or it faked a coup attempt in order to justify a crack-down, it’s hard to know for sure), and consolidate power. Then, this year there was an election which the President handily won and there was suppression of opposition, so there was a further move towards an authoritarian government. This pushed valuations down further.
Then, President Trump applied strong rhetoric and punitive tariffs in order to punish the regime for the imprisonment of an American missionary and to push for his release. This has added a rout on the Turkish lira on top of the rout from the political instability stemming from the election.
All of these things pushed valuations down to highly compressed levels. Between the beginning of the year and mid-year, prices of the Turkish equities fell almost a fifth in value, while lagging earnings rose almost a tenth and forward earnings rose almost a seventh. This made valuation a very attractive proposition at mid-year.
In addition, in general high bond yields, compared to the rest of the world, are generally historically a positive thing. They tend to show that there is a lot of the emotion of fear already expressed in markets, and they mean that there is a pressure for currency to rise as investors arbitrage from lower yields to higher yields. In addition, high yields tend to correlate with yields moving down in the future as things move back to normal. Turkey had the highest bond yields in our investible universe as of mid-year at almost 17%.
Also, yields had gone up from the beginning of the year to mid-year. Recent large increases in interest rates tend to correlate with future good returns, normally, because of the same idea of returning to normal. So, all other things being equal, having low valuations and high bond yields which just went up in the last period is a sweet spot for good returns, if things are normal.
But things aren’t normal in Turkey. The biggest problem has been political risk. Sometimes high bond yields and low stock valuations and other signs of fear mean a bargain basement pricing. But sometimes the fear is justified -- for example, when a nation is politically unstable. The issue of political risk deserves a separate treatment all of its own, which I'll outline in a future article in the series. Suffice it to say that the trick is not just in spotting politically risky countries, but rather in spotting which countries to underweight so that the portfolio is ahead of the game. Simply screening out risky countries historically means throwing out more big up years than big down years.
There was a strong case for Turkey based on valuation, but there were also strong cases against it based on political instability and on foreign debt exposure, which I’ll talk about in future columns.