Kent Peterson

Since the end of January, the US Dollar, on a trade-weighted basis, has risen 5.4%.  Historically, when you look from 1975 onward, that is a lot.  Percentile-wise that puts the current move in the top 16% of all moves since 1975.  And when you consider upside standard deviation alone, its about a 1.82 Z move, meaning it’s a pretty rare, though not entirely an outlier, event.

So what happens economically and financially when your currency goes up a lot?  If you are an island economy or very export dependent like Japan, a large upside move in your currency should be painful, because it makes your exports uncompetitive and increases your import costs.  Or………so the conventional wisdom goes.

When you look at what actually happens to domestic markets in the developed world historically, as well as what happens to industrial production when the trade-weighted currency of an economy rises 5% or more, you find two things.  First, a big jump in the currency really doesn’t impact major financial markets at all.  Stocks and bonds rise normal amounts, perhaps because the FX move is seen as a temporary effect.  By contrast, a rise in the currency does appear to actually dampen industrial production, which is what one would expect.  Put another way, large currency moves in the developed world appear to have important economic consequences, but little impact on financial markets, perhaps because the economic impact is so short-lived.

Here is what happens to industrial production in the US and among 17 other developed countries after the trade-weighted currency rises 5% or more, using data from 1975 onward.  Historically there are 87 months when the US currency was up more than 5% over a 5 month stretch, and 769 months for the other 17 developed currencies.  As you can see in the chart, US and other developed markets both experience a period of about 5 months where industrial production growth contracts, on average, modestly.

By contrast, US and other developed fixed income and equity markets are pretty much unfazed by large currency appreciations, as you can see in the next two charts.

First Bonds.

Now stocks.

Obviously a few countries are exceptional.  Australia, Norway and Hong Kong do see equity declines for several months after a large swing upward in the currency, but with the exception of Norway, every equity market is higher a year later after a currency jump.

So the next time someone tells you the large rise in the currency of a developed nation is going to make them uncompetitive and cause their markets to fall, you should probably ignore them, and go have a coffee.

Please follow and like us:

Consumption growth matters a great deal in the developed world.  It’s about 2/3^rd of US GDP.  So you would think there would be a pretty clear linkage between higher consumer growth rates, faster earnings growth and superior stock performance.   Think again.

In fact, it’s quite the opposite.

When you look at periods when annual retail sales growth was in the top decile and bottom decile over the last 10 years, and then measure subsequent stock returns, the bottom decile wins quite handily.  The chart below shows an 18 developed country average since 1960 of stock index returns after annual retail sales growth was in the bottom or top decile over the prior decade.

We have some sample size here too.  Between the 18 countries, there were 570 months when annual retail sales was in the top decile and 856 months when it was in the bottom decile.  To prevent one country dominating, I simply took an equal average across all countries.

What the chart shows is not a subtle difference.  When retail sales growth is in the bottom decile, you make, on average, just short of 25% over the next 2 years.  Buy when its in the top decile, and you make a mere 1.43% over the next 24 months.

Why?

I don’t know but there are two probable explanations, both of which may be at play.  Weaker growth leads to lower interest rates which begets borrowing and expansion.  Higher growth leads to higher rates and things slow down.  Markets are also discounting mechanisms and people aren’t real good at imagining a world different than the world they currently inhabit.  So when consumption is weak, people and markets are overly pessimistic.  Similarly, “happy days are here again” consumption is probably the period the market has become overly optimistic in its future pricing.

At present, only one country, France, sits in the top decile of 10 year retail sales growth.  No one, unsurprisingly, is currently in the bottom decile.

Please follow and like us:

Last week we looked back through US and several developed countries financial history to see what effect of a 1% annual gain in short term interest rates had upon subsequent equity returns. The results showed that US equities have a fairly negative outlook over the next 4-6 months from an interest rate history perspective. This week we are going to be applying the same test to see the effect on long bond returns. Unfortunately, the news is also bearish for expected long bond returns for the next 4-6 months.

As noted last week, 3-month T-Bills have experienced annual changes of 1% or greater 21 times since 1960, while 2-year notes have seen 28 such increases. The first chart below shows what has happened to US long bond returns (meaning Treasuries of around 20 year maturities) over the next year. Like stocks, negative returns seem to bottom around 4-6 months out, with losses of 1.5% to 2% being highly probable if the past is any guide. Though 1 year later, the pain is over and you are back to flat. Whether US bond investors should expect this bond bounce back to repeat itself this time remains a question mark for me, however. Keep in mind that interest rates and bond yields have been in secular decline since their peak in the early 1980s, and there are a lot of tells of future inflation looming on the horizon at present.

Extending the same analysis more globally to 8 developing major countries, we can see a similar pattern, though with more sample size. Internationally, 3-month T-Bills have shown annual changes of 1% or greater 114 times since 1960, while 2-year notes have experienced 121 increases. Like in the US, the long bond’s maximum losses happen about 5-6 months out, but it’s worth pointing out that the loss is a much more modest 25-50 basis point decline. And also like the US, returns 1 year later are flat and even up.

We will conclude the same way we did last week, by pointing out that while the US and Canada have recently experienced 1% annual increases in short-term rates, the rest of developed world has not. So investing in other developing bond markets outside North America remains a very attractive relative value proposition for the next 4-6 months.

Please follow and like us:

With the exception of a few anomalous periods, consumer credit has pretty much always grown (as you can see in the first chart below).  Not shockingly, it actually contracted briefly during the crisis of 2008, as households cut back.

When you look at longer-term growth rates in consumer credit, things look fairly normal.  We are not near historic 5 year growth rate lows of 10%, like we were in the crisis, or even the 20% lows we saw in the early 1990s.  Consumer credit 5 year growth is hanging in there around 30%, which is historically a little on the low side, but not terribly so.

Dig a bit deeper beneath the surface, however, and you do see two interesting recent trends.

First, the government has become a major player in consumer credit markets, especially since the financial crisis when risk averse financial institutions became much stingier.  Below is a chart of the composition of who is lending to consumers by institution type.  Banks have always been, and continue to be, the largest providers of consumer credit.  Historically, they have provided 50-60% of consumer funds, though that has dipped closer to 40% since the mid 90s.  Financial companies and credit unions show some variability through time, with financial companies seeing a 10% decline in their share of the lending market since 2010.  But by far the largest change, and the one most worrying if you are a believer in free markets, is the huge jump in the presence of the government in providing credit:  the government is now close to 30% of the consumer credit market now vs. 5% ten years ago.  This is almost entirely a result a second secular trend in the credit markets—the explosion of student debt.

If you look at a longer-term historic chart that breaks down what types of consumer loans were taken, you don’t get a ton of insight because student loans have only been broken out by the Fed since 2006.

What’s interesting is how credit card debt was fairly non existent in the early 1970s at a mere 5% of consumer debt and ran up to 40% in the late 90% and has since come back down closer to 25%.  Car loans, by contrast, have been pretty steady at 30% of total consumer credit.

It’s only when you look at a breakdown of consumer credit since 2006 (when the student loan data becomes fully available) that the changes in US consumer debt composition become so obvious.  As you can see in grey line in the chart below, student loans have moved from 20% of all consumer credit in 2006 to close to 40% now to become the largest percentage of all consumer credit.

The government’s increasing market share of consumer credit is directly tied to the rapid growth in student loans.

The problem is that, in the long run, the growth rate in student loans is utterly unsustainable.  There is too large a gap between the higher growth rates in student loans and lower growth rates in household incomes.  As we learnt from the last crisis, there is only so long that debt growth can outrun income and asset growth.  If borrowing growth exceeds income growth, debt payments either cannot be made and we get defaults, assets have to be liquidated to fund debt payments (young people out of school however have no assets), or students will have to go to family to fund their debt.  None of this is good for consumption or the growth in earnings of companies are tied to younger consumers.

At some point either student loans growth will have to slow, or these loans will have to be written down.  This may or may not be a “credit bubble” since bubbles typically require some leverable asset that increases in value, but nomenclature aside, at some point there will be reckoning.

Please follow and like us:

A slightly different version of this post, written with Chris Shuba and Joe Mallen of Helios Quantitative appeared here: IRIS

You are a US investor that wants to deploy some of your money overseas.  You think international developed market stocks are attractive relative to US stocks, and you also think the US $ will decline over the period you intend to hold your investment.  Your investment decision is logical to you. But you have choices:  You could a) simply invest in a traditional index like the MSCI EAFE, b) invest in a fund that systematically emphasizes a particular factor like a value fund that only buys cheap overseas stocks, or c) invest in a developed fund that blends several factors together, like the JPMorgan Diversified Return International Equity ETF (JPIN).  What is the best choice?

Investing in a traditional international market capitalization index like the MSCI EAFE is not a bad choice. It has delivered nice returns for a US investor, especially uncorrelated outperformance in the 1970s and 1980s, and helped to diversify a US-only portfolio.

Your second choice is to invest in one particular investing style because it makes sense to you.  You might believe that a fund or index that chooses the cheapest or most attractively valued stocks based on metrics like Price to Earnings (PE) is best.

You could go find a discretionary portfolio manager who only buys stocks he deems to be cheap.  Typically the concept of “cheap” is based on some absolute metric that the manager has in mind, such as never buying a stock with a PE greater than 15.  If there are not enough stocks that are attractive, he will hold his money in cash until he finds the prudent bargains he seeks.  This prudence also obviously risks possible underperformance from being absent from the market.

The alternative is to buy a value index or fund that systematically only buys the cheapest stocks in a particular investment universe.  So if there are 1000 investable stocks available, the index ONLY buys the cheapest decile of 100 stocks and is always fully invested in the 100 securities that are relatively cheapest.  This is an investment approach that a discretionary manger may disdain.  The discretionary value manager may look at those same 100 stocks and think they are pricey.  But nevertheless, academic research has shown that always being fully invested in the relatively cheapest percentiles of stocks in the US has produced superior returns over many decades.

Such a portfolio is called a “factor” portfolio.  Why the name?   In the early 1960s, academics introduced the concept of beta and demonstrated that individual US stocks had sensitivities to, and were driven by, movements in the broad market.  In the early 1990s, academic research began to show that other “factors” such as value and size also drove US stock returns.  Since then, several factors have been identified as driving individual stock outperformance: value, size, volatility, momentum and quality.  Stocks that are cheaper, smaller, less volatile, have more positive annual returns and higher profitability have historically outperformed their peers.  It turns out these factors also work internationally.

Of all the factors, value is the factor that has been the best known the longest (even before it was academically identified as a “factor”), thanks to the books of Warren Buffet’s teacher Ben Graham.   And if you look abroad at an array of developed global markets and create a value index and compare it to its simple market capitalization weighted brother, the historic outperformance of value has been stunning.  Until recently.

While there was some variability by country, on average from the mid-1970s up until 2005 a value factor portfolio in a developed market outperformed its market cap weighted index by about 2% a year.  That’s a lot. By contrast, since 2005, the average developed country value portfolio has underperformed a market cap indexes by about -40 basis points.  Which is the danger of investing in one factor.  It may not always work at every point in time.

So if investing in one factor like value runs the risk of underperforming, how about a multi-factor international developed equity portfolio?

Below is a breakdown of individual factor portfolios’ performance in international developed equity markets since 2005, an equal weighted factor portfolio as well the performance of the MSCI EAFE as our performance reference.  Note that, for the last 13 years, value has been the poorest factor by far, while the others have handily beaten the EAFE.  An equal weighted portfolio of all 5 factors, while not as optimal as some of the individual factor results, beats the EAFE by 1.6% and has an information ratio, or risk adjusted returns that are superior by 37%.  The equal weighted factor portfolio also has the advantage of not having to predict which factor will work when, so even when a factor like value does not beat the market, the other factors can pick up the slack.

The equal weighted factor portfolio has one other advantage over the market cap weighted alternative. Note in the chart below how well the portfolio outperformed in the 2008 crisis, so it tends to do relatively well in highly volatile sell offs.

While it’s not inconceivable that one or two of these factors could erode, or underperform for a stretch, the fact that you have exposure to multiple factors in a portfolio that seems to do especially well in crises suggest the multi-factor blended portfolio remains the most attractive way to invest in developed markets.

Please follow and like us: