Wednesday, September 13th, 2017

Untangling the Currency Effect

Few would dispute the idea that the US sets the tone for what happens in global financial markets, but occasionally it is worth surveying how the choice between fixed income and equities looks to investors in other parts of the world. This is particularly true when FX markets are volatile or trending strongly.

This year, US investors have enjoyed significant windfall returns from international equities thanks to the weak dollar. However, when viewed from Europe, the outlook is not nearly as sunny. Our euro-denominated model is down to 40% equity exposure, having started the year at over 90%. The strong euro means that returns from US equities are negative YTD, while returns from Japan and the UK are only just above zero, and certainly well below their cost of risk. We don’t expect this to change while the weak dollar regime lasts.

The question is what happens next. We could move into a dollar recovery phase, but it is also possible that we move into a more stable FX regime, in which case investors all-round the world will have to stop fretting about currencies and start to prioritise a domestic view. We can’t isolate local equity and fixed income markets from the currency markets, but we can look at models which restrict the choice of assets to domestic equities and domestic government bonds all based on local currency returns.

We track these local for local portfolios on a weekly basis for over 40 countries and we have a 40-page chart book which shows the make-up and performance of our recommended portfolios, as well as our forward-looking indicators as to how we think the portfolio will evolve. The charts on the next page are the two summary charts, which look at the major developed and emerging markets (with the Eurozone regarded as one “country”).

The first is backward-looking and plots the current equity / bond split on the x-axis against the change in the split over the last four weeks. The second plots the current split against our forward-looking indicator. In both charts the top-right quadrant is the most desirable, representing a high and rising allocation to equities (which in turn is based on an improvement in their risk-adjusted returns over the period). In each chart, we plot the last four data points for each country so that we can see the speed and direction of the trend and whether it is accelerating or not.

The key messages are as follows.

China has the highest allocation to equities at the expense of government bonds. The four-week change is zero, because we have been at or close to 100% equity since late June. The forward-looking indicator is also zero, implying that there is no risk of the model reducing its equity exposure in the short term. From a real-world perspective, this is a conclusion that we would be happy to endorse until at least the 19th Party Congress which happens in mid-October.

At the other end of the table, Russia has 2% exposure to equity and a forward-looking indicator of only 4%, suggesting that the highest exposure we could hope over the next month would be 6%. However just because the indicator is positive, it doesn’t mean that the equity exposure will actually rise. Russia had a higher indicator in late July, but failed to make any progress. A view on Russia, is of course heavily dependent on a view on oil and we don’t see any signs of a big recovery in any of our commodity or equity sector models.

The recent strength of China allows Brazil, Australia and Canada to make good progress. The commodity link is obvious and the fact that all three have responded in the same way is comforting. Because of domestic political events, Brazil is much the strongest of these three, with an accelerating trend and an improving forward indicator. It is the only country which is in the top right quadrant in both charts.

By contrast, the Eurozone is one of two countries which is in the bottom left quadrant of both charts; the other is Mexico. The Eurozone has moved from a 65% equity exposure to 35% in a period of just four weeks and the rate of decline has accelerated slightly. The forward-looking indicator has suffered an equivalent decline, but the rate of change is decelerating as it starts to approach the zero-limit. This is the same pattern that we observe in the euro-denominated portfolio which is allowed to invest in international equities, but it makes the point that domestic equities are also much less attractive to euro investors than they were in late July. It is also no surprise to see that that Switzerland is following exactly the same trajectory as the Eurozone, but with a three-week time lag.

So, what about the other big developed markets, Japan, UK and USA? Four weeks ago, all three were on the boundary between top-right and bottom-right in the current situation vs past trend chart. i.e. they all had equity exposure ranging from 80-90% and had been stable at that level for the previous four weeks. Now they are firmly in the bottom right quadrant: high but falling equity exposure and all of them show an accelerating rate of decline. The forward-looking indicators for Japan and the US are still mildly positive, but the UK has turned mildly negative..

This suggests that we are now at the point where declining risk-appetite in Europe has started to have an impact on the USA. The trouble with this formulation is that it requires Europe to be causing a problem for the US, which immediately requires investors to specify a view on exchange rates. We prefer to frame this as a statement about negative feed-back loops affecting investors across a broad range of developed markets in the wake of a significant cumulative FX shock.

Perhaps the best way to sum this up is to use a cap-weighted proxy for the World. This shows that the average recommended equity allocation has fallen from just under 90% at the end of July to its current reading of 65%. The forward-looking indicator is mildly positive at the moment, but at a level which could easily be reversed by one bad week.

None of this changes our view that a dollar-based investor ought to have a high exposure to international equities, while the weak dollar regime persists, and that a euro-based investor ought to favour fixed income. From this it follows they should do the opposite if we move into a dollar recovery phase. However, even if the dollar and the other major currencies were all to stabilize at their current levels, the level of risk-appetite across the world and in the US would likely be significantly lower than it was in the summer. The only real exception to this China, where the government manages the exchange rate and most of the other financial markets.

Leave a Reply

Wednesday, September 6th, 2017

The Probability of Loss

This note follows on from last’s weeks discussion of the opportunity cost of being ready for a bear market. We spend a lot of time talking about risk, which we normally define as the volatility of weekly total returns. However, we also look at other definitions, such as the probability of loss; i.e. the chance of experiencing a negative return over the next 52 weeks based on the annualised return and volatility of the last 52 weeks. This is, of course, the view from the rear-view mirror, but the current set of numbers is very interesting and helps to explain why so many US investors are reluctant to abandon the bull market in US equities.

Based on the last 52 weeks of data, the chance of US equities delivering a negative return over the next 52 weeks is just 3%. This compares with a median observation of 17% over the last 22 years. It’s what you would expect at the end of a long bull run, but it’s very hard to get out when the odds are so favourable. The forward-looking bear scenario has to be extremely convincing, and if it’s so convincing you have to ask why other investors don’t see it.

For a dollar-based investor, the probability of incurring a loss in other equity regions is also very low: 4% for Japan and Emerging Markets; 3% for the Eurozone and 16% for the UK – up from 5% a month ago. These numbers are clearly a function of the weak dollar, which many investors expect to reverse in the near future. However, there is a growing number of currency strategists talking about PPI valuations (where the dollar is still overvalued) rather than interest rate differentials. If dollar weakness continues international equities could stay attractive to US investors for a long time.

Contrast this with the fixed income market. The probability of loss for medium-dated Treasuries is 62%; for Investment Grade, it is 37% and 40% for dollar-denominated EM Bonds. Only High Yield offers the same low probability of loss – 3% – as US equities. Its running return is actually higher than investment grade, while its recent volatility is lower. All fixed income assets apart from High Yield have a current probability of loss which is well above their 22-year median. So, when US investors are invited to de-risk their portfolios, they are being asked to swap into assets which mostly have a higher probability of loss in absolute terms and where the probability of loss is elevated by comparison with history. One day this advice will be right, but for the moment it looks premature, absent a geo-political crisis.

For non-US investors, the situation is very different. For Eurozone investors, the probability of loss from Eurozone equities is now 13%, up from 3% at the beginning of July. For international equities, the numbers range from 38% for the UK down to 13% for Emerging Markets. Apart from EM, all these numbers are significantly higher than they were two months ago. For German Bunds, it is 66% and 52% for euro-investment grade. Based on the way the base effects work, these numbers are likely to carry on rising for all equity regions and to decline gently for euro-denominated fixed income.

Furthermore, the size of the loss from fixed income is unlikely to be material. The probability of a 10% loss in Bunds is only 4%. Eurozone investors may soon have to choose between the high probability of a small loss in euro fixed income or a rising probability of a bigger loss in equities. For the moment, US investors have a much happier choice.

Leave a Reply

Wednesday, August 23rd, 2017

Winter Is Coming

With apologies to all fans of A Game of Thrones, and those who have yet to become addicted, the problem with motto like the title is that it is very easy to sound like a perma-bear. In the TV show we were introduced to the theme in the first episode of series 1. We are now at the end of series 7 and it has only just arrived. If the show had only been about the onset of winter, the producers would not have had enough material to fill two series, let alone seven. In entertainment, as in investment, it’s what you do between bear markets that really matters. The opportunity cost of always being ready for winter can be huge.

As we return from this year’s holiday period, our US asset allocation model is resolutely pro-equity, though there is no shortage of threats which could rapidly alter this position. Our non-exhaustive danger list is probably no different to anyone else’s: a property crash / credit crunch in China, failure by the Trump administration to advance a significant tax reform package, a Minsky moment regarding equity valuations, the shutdown of the US Federal government, a sudden surge in FX volatility or (god forbid) a nuclear strike emanating from North Korea. However, the onset of winter requires that one of these risks actually happens AND that the future probability of other bad things happening also increases.

The second part of this definition is more important that than the first. Without some form of contagion, investors’ response to one of these risks crystallising will likely resemble the current “buy-the-dip” consensus. The quickest way to understand this is a simple “what-if” exercise. Let us assume that US equities fall by 10% between now and the end of September. The trailing 52-week return declines from 13.4% to 3.1%, and realised volatility rises from 8% to about 10%, which is still well below the long-run average. The 52-week Sharpe ratio falls from 1.64 to 0.32, which is only just below the average for the last 10 years.

There is nothing very frightening in these numbers. We need a correction of 15% in US equities to drive the Sharpe ratio seriously below zero and even that would not be enough to shock them out of their low volatility regime – although the medium-term consequences might be. But this is precisely our point. We need two or more risk-scenarios to go wrong either simultaneously or very close together. Without contagion, the chances a cycle-ending bear market are really quite small.

Contagion also operates between countries, there needs to be a credible transmission mechanism and other markets need to be susceptible to infection. Based on current data, the US only has the 10th best 52-week Sharpe ratio in local currency terms. Other countries tend to have higher volatility but they also have higher running returns and their current rallies started later, so they don’t look as tired. Using the same 10% correction scenario for all countries, the US drops to #13 in our list of 25 major markets. The Eurozone and EM Equities as a whole would both have better risk-adjusted returns. So too would individual countries like Japan, Korea, China and Taiwan in NE Asia and France, Spain, Italy and the Netherlands.

In our view, global equity markets are likely to experience a correction this Autumn, but one snowstorm does not make a winter. Investors will survive it. It’s what happens afterwards that matters.

Leave a Reply

Thursday, July 27th, 2017


Our models have a split attitude towards Eurozone equities. Our dollar-based global equity model has all the individual countries at or close to the top of the leader-board for all the usual reasons: above-trend and accelerating GDP growth, strong currency, reduction in political risk, safer banking system. This is in obvious contrast with other developed markets like Japan, the UK and USA which have weak currencies and political gridlock.

However, our euro-based asset allocation model has reduced exposure to Eurozone equities by 9 percentage points in the last 5 weeks. This is part of a general move out of equities into fixed income. This sounds strange since our fixed income portfolio is currently producing negative returns. However, when we compare the change in the running returns over the last two months the reason becomes clear. At the end of May, our global equity portfolio, which is heavily overweight Europe and underweight the US had a running return of 20%, compared with a 0% running return from our fixed income portfolio (with all constituents close to neutral weight). As of last week, the numbers are -5% for fixed income, but only 3% for equities. The gap between the run rates has closed from 20% to 8% in less than two months, and the composition of the equity or fixed income portfolios has no more than a marginal effect. Our two big exposures relative to benchmark are the consensus overweight on the eurozone and the underweight on the US.

The elephant in the room is of course the strength of the euro, which has risen from 1.059 against the dollar at the beginning of April to 1.166 as of last week. The good news is that it now looks overbought in the short term against the dollar and the yen. The bad news is that is still below the middle of its three-year range against the dollar (approx. 1.20) and is trending higher against all major currencies. We have no problem with the idea of a short-term pull-back, but all the reasons which make eurozone equities so attractive to international investors also mean that that euro is likely to appreciate over the medium term.

Aside from the usual sector and country rotation, we see three principal effects if the euro trends higher over the rest of this year. First, euro-based investors would gradually shift out of international equities, because the returns in euro terms would not be as attractive as they were. Second, Eurozone companies would start to downgrade earnings growth expectations for 2018, mainly because of translation effects. Third, these two factors lead would to a significant shift away from equity towards fixed income. The strength of the euro leads to tighter monetary conditions in Europe, allowing the ECB to postpone the day when it has to contemplate a sustained rise in interest rates or start to shrinking its balance sheet.

Of course, there are many reasons why euro strength may not continue. It’s possible that the result of the German and Italian elections may not be so market-friendly as the French and Dutch elections. But that would hardly be good for Eurozone equities either. We still think they will be the largest single position in our asset allocation model for some time to come, but they will be a normal overweight in a portfolio which is modestly overweight equities. As we argued shortly after M. Macron’s election, we are already past peak Euro-phoria.

Leave a Reply

    Full Blog Archive:

  • Search Blogs by Category