by Asset Allocation Committee

Due to the upcoming holidays, the next edition of this report will be published on January 6, 2017.

The Fed gave us a modest hawkish surprise last week, calling for three rate hikes in 2017 rather than two.  The news has boosted Treasury yields and lifted the dollar.  Equities mostly absorbed the news without incident.

Here is a chart of the FOMC’s average dots over the past two years.

The fuchsia dots represent the most recent meeting.  The dots have stopped their steady progression toward lower levels.  For better or worse, the path of policy expectations from the dots suggests that the FOMC is becoming comfortable with this path of hikes.

Here is the dots plot from September.

Note the purple line, which is the LIBOR-OIS curve from the meeting day.  It has jumped from where it was on the meeting date in September, shown by the red line.  For the past few years, the FOMC dots have tended to decline toward the market.  The rise in the LIBOR-OIS curve suggests that process is reversing.

This is the new dots plot, released at the December meeting.

For 2017, the median forecast is currently 1.375%, up from 1.125% in September.  For 2018, the median is up to 2.125% from 1.875%.  Two participants see no change next year but one of those is probably St. Louis FRB President Bullard, who has decided not to participate in the dots procedure.  Although market expectations continue to lag, we did see the LIBOR-OIS rate rise to 1.25% from 0.875% in September.

This can be seen in the deferred Eurodollar futures.

The jump in yields since Trump’s election has been striking.  We are approaching the highest level of implied rates since the “taper tantrum.”  This rise triggered the onset of the dollar rally in mid-2014 and we note that the dollar has been rising since the election.

To get some sense of where policy is in relation to the neutral rate, we use the Mankiw rule model, incorporating the recent rate changes by the FOMC.  This model attempts to determine the neutral rate for fed funds, which is a rate that is neither accommodative nor stimulative.  Mankiw’s model is a variation of the Taylor Rule.  The latter measures the neutral rate using core CPI and the difference between GDP and potential GDP, which is an estimate of slack in the economy.  Potential GDP cannot be directly observed, only estimated.  To overcome this problem, Mankiw used the unemployment rate as a proxy for economic slack.  We have created four versions of the rule, one that follows the original construction by using the unemployment rate as a measure of slack, a second that uses the employment/population ratio, a third using involuntary part-time workers as a percentage of the total labor force and a fourth using yearly wage growth for non-supervisory workers.

Using the unemployment rate, the neutral rate is now 3.63%.  Using the employment/population ratio, the neutral rate is 1.08%.  Using involuntary part-time employment, the neutral rate is 2.79%.  And finally, the wage growth model puts the neutral rate at 1.56%.

It is still uncertain which of these variants best reflects slack (or lack thereof) in the economy.  Although we tend to think that wage growth or the employment/population ratio is the best measure of slack, the key is which one policymakers view as the most consistent with measuring slack.  At this point, we don’t know, although we think the hawks are probably relying on the unemployment rate variant while the chair and most of the doves probably believe the involuntary part-time employment variant is the best measure.  The involuntary part-time employment variant is most consistent with six rate hikes over the next 24 months.  That path would bring the policy rate near neutral; however, if they are wrong and, for example, the employment/population ratio is actually correct, then policy will be overly restrictive (assuming that ratio doesn’t improve dramatically).  Thus, the FOMC is moving rates higher in a slow fashion to allow them time to adjust if it turns out there is more slack (reflected by the lower neutral rate variants) than some data would suggest.  Of course, by going slow, assuming the higher neutral rate variants are correct, the Fed could keep policy overly accommodative longer than it should.  However, as long as the economy remains globalized and deregulated enough to allow for the nearly unfettered introduction of new technology, being late isn’t all that risky.  That assumption would change if the incoming President Trump puts up trade barriers.  Thus, the path of monetary policy could be a risk factor in the upcoming year.

View the PDF

by Asset Allocation Committee

Equity markets are expensive by numerous measures and have become even more extended in the wake of the “Trump Rally.”  As noted in our weekly P/E update (found in the last section of the Daily Comment), our four-quarter measure of the P/E is elevated.  Another well-known derivation of the “Buffet Indicator” has also reached lofty levels.

This ratio divides the market value of equities outstanding from the Federal Reserve’s Financial Accounts of the U.S. by GDP.  Although the ratio is off its recent highs, it is still at the upper end of its historical range.

At the same time, the optimism isn’t unfounded.  Proposals for corporate tax reform could bring repatriation of the more than $2.0 trillion of corporate cash parked overseas.  Tax cuts would boost after-tax earnings.  The “E” part of the P/E could rise, making the market less expensive.  Another factor to note is that investor cash levels remain elevated.

This chart shows the level of retail money market funds relative to the S&P 500.  Since 2011, equities have tended to trend higher when money market fund levels exceed $920 bn.  Once money market holdings fall below that level, equities have tended to stall.  It appears that investors were building cash in front of the election but, so far, retail investors are not aggressively entering the equity markets.  The weekly mutual fund flows suggest that investors are selling out of bonds which probably accounts for the recent rise in money market funds; to date, there isn’t any evidence in the same data that equity mutual fund buying is increasing.  This data would suggest that there is ample fuel for further gains in equities.

So, if equities are rich but could work higher, how high is high?

On this chart, we log-transform the weekly close of the S&P 500 starting in 1928.  The red line is the long-term trend.  We have regressed the trend line against the weekly close and have calculated the standard deviation at the ½ level, one deviation and two deviations from trend.  Only twice in the past 88 years has the index exceeded two standard deviations to the upside and it has only touched that level on the downside three times.  We are currently below ½ deviation on the upside, suggesting that, on a long-term trend basis, equities are not all that extended.

If we reach +½ standard deviation by mid-2017, the S&P 500 would reach 2480.70; the year-end reading would be 2562.22 by the same measure.  By comparison, reading the +one standard deviation line would put the index at 2995.80 by June 2017, and 3094.24 by year’s end.  We are not anticipating a rise to +one standard deviation next year but a rise to the +½ level isn’t out of the question.

We still don’t know what the specific policies will be of the incoming Trump administration.  Tax policy is complicated and there is always the potential for mistakes.  The current euphoria surrounding other policies will probably fade over time.  And, although the above analysis suggests that equity markets are sporting high valuations, cash levels are high and current equity levels, while above trend, haven’t reached levels of concern.  Thus, for the time being, the path of least resistance for equities is probably higher.

View the PDF

by Asset Allocation Committee

The rapid rise in longer duration Treasury yields since the presidential election has been surprising.  As of December 8, the 10-year T-note yield was approximately 2.40%.  Although President-elect Trump’s policies will probably be inflationary, it is still unclear how much of his arguably vague plans will get passed.  It is possible the FOMC will become more hawkish and we have seen some increase in rate hike expectations.[1]  Still, our 10-year T-note model is putting the fair value yield at 1.85%.  Assuming fed funds at 1.25% still only raises the 10-year rate to 2.20%.  Taking oil to $60 and assuming the 1.25% fed funds only raises the fair value yield to 2.27%.  Only when assuming steady oil, fed funds at 1.25% and German bunds at 1.25% (up from the current 33 bps) does the yield even reach 2.40%.  The current spike in yields can be best justified by assuming a significant jump in inflation expectations.

In our yield model we use the 15-year average of CPI as a proxy for inflation expectations.  This assumption comes from the work of Milton Friedman, who postulated that inflation expectations are derived over a long-term time frame.  We realize our calculation is a proxy but have refrained from using more market-based expectations because of their lack of predictability.  If one assumes that nominal rates are the sum of the expectations of real rates plus inflation forecasts, inflation forecasts are very important to predicting nominal interest rates.

If the lifetime experience of inflation is important, then what is the most important age?  We estimate that 60 is a reasonable age; the average age of the Senate is 61 years, the current FOMC average is 62 and the average age of an S&P 500 CEO is 57.[2]  Simply put, it’s around the age of 60 that people come into power in politics and business.  We believe that their personal experiences color the expectations of any investor and so using 60 as an influential age makes sense.

This chart shows the adult experience of inflation for a person turning 60 from 1932 to the present.  To reflect the adult experience, we use the average annual change in CPI from ages 16 to 60.  Note that inflation experience rose into the late 1940s and stabilized into 1960, when it fell sharply.  This was the generation that entered adulthood during the Great Depression.  It is interesting to note that as rates began to rise in the mid-1960s, the inflation experience steadily rose as well.  Essentially, the rise in rates coincided with the rise in inflation experience.  However, after peaking in 1981, bond yields began a steady drop into the current year despite the relatively high level of inflation experience.  On the other hand, T-note yields exceeded the inflation experience of 60-year-olds in absolute terms until 2002.

This chart shows the actual inflation rate compared to an average 60-year-old’s adult experience of inflation.  In general, bull markets in bonds tend to occur when the actual inflation rate is persistently below the average rate.  Bear markets happen when the opposite condition is in place.  Currently, the actual inflation rate is still well below the average rate, suggesting that the bull market in bonds should have more time to run.  However, our worry is that the average 60-year-old is unusually sensitive to inflation fears and thus may overreact to the incoming president’s policies.  In other words, inflation expectations may become unanchored rather quickly, forcing the Federal Reserve to turn unexpectedly hawkish.  Thus, we are taking a more cautious stance on fixed income into 2017, expecting higher yields and greater duration risk.  At the same time, we will be closely monitoring the economy in light of less accommodative monetary policy.  Most recessions occur because the Fed tightens too much.  We don’t expect that to become a problem until late next year or early 2018 if the Fed continues to raise rates.  So, for the upcoming year, we expect a weak fixed income environment.

View the PDF

_______________________

[1] For example, the two-year deferred Eurodollar futures, which measure three-month LIBOR two years into the future, have jumped nearly 50 bps since the election.

[2] http://fortune.com/2015/12/13/oldest-ceos-fortune-500/

by Asset Allocation Committee

Last week, we discussed the likely implications of President-elect Trump’s policies on the debt markets.  This week, we will look at the impact on the dollar.  Since the election, the dollar has generally moved higher.

Using the Bloomberg dollar index, a broad-based currency measure, the dollar rose nearly 5% after the election.  There are a number of arguments behind the rise.  Trump campaigned for fiscal expansion, which could include both infrastructure spending and tax cuts.  The expected fiscal expansion could lead to tighter monetary policy and this particular combination is usually thought to be bullish for the dollar.

This box describes the expected outcomes from the interplay of fiscal and monetary policy.  This is a rough guide; the actual outcomes are mostly driven by the degree of policy adjustment.  In the early 1980s, the combination of real fed funds of nearly 8.5% and a fiscal deficit of almost 6% of GDP led to a very strong dollar (the “Volcker dollar”).  Market behavior may be anticipating a repeat of this outcome.

However, this assumption depends on the FOMC moving to tighter policy, almost a “hard money” stance of the Volcker years.  Simply put, we don’t know for sure whether this will be the outcome.  We believe that there is a political struggle in the Trump administration between the GOP establishment and right-wing populists.  To personalize the sides, we view it as Speaker Paul Ryan versus Steve Bannon.  The FOMC has two open governor positions.  If Ryan’s wing of the party wins, we will likely see the Fed change into a hard money central bank, which is a quadrant two outcome on the above table.  On the other hand, if Bannon’s wing wins, it is possible that we will see doves appointed to the Federal Reserve.  That scenario could lead to a quadrant four outcome, which would be quite different from what the market expects.

The policy situation isn’t the only supporting factor for a stronger dollar.  It is estimated that over $2.0 trillion is held by U.S. companies offshore in order to avoid corporate taxes.  If corporate taxes are reformed, at least some of this money will come back home which would lift the dollar.  If Trump were to put up trade barriers as promised, the current account deficit would shrink, which would reduce the supply of dollars and boost the dollar’s value as well.  Thus, for now, we expect the dollar to get the benefit of the doubt and it will likely continue to appreciate.

What is hurt by a stronger dollar?  The two asset classes most at risk from dollar strength are commodities and emerging markets.  Since the election, commodity prices have been mixed; the Bloomberg commodity index is actually higher since the election, up 2.7%.  Industrial metals are up 6.4% over this time period and energy is up 5.9%.  At the same time, gold is down 7.4%.  The MSCI Emerging Market Index is down 2.8% since the election.  If the dollar continues to appreciate, these asset classes will likely face further declines.

View the PDF

by Asset Allocation Committee

Trumponomics looks as if it will be a combination of fiscal stimulus, trade restrictions and deregulation.  It looks very likely that environmental regulations will be reversed and there have been promises of financial deregulation as well.  The first two will likely reflate the economy.  Proposed deregulation may help hold down energy prices but financial services are not a major contributor to inflation (only about 0.24% in CPI) anyway.

With reflation on the horizon, we have seen a rise in the 10-year yield.  Even though we would expect a retreat in yield during the next recession, it is likely that the secular bond bull market that began in the early 1980s is coming to a close.

The chart above shows the 10-year T-note yield from 1921.  Perhaps the most important issue to remember is that when the last secular bear market began after the lows were made in 1945, the next peak took 36 years.  It took eight years before yields doubled.  Although the regulatory environment is different, it takes a while for bond yields to reach really high levels.  Still, the tailwind for financial assets that this bull market represents is noteworthy.

This chart shows the 10-year T-note yield and the cyclically adjusted price/earnings ratio (CAPE) that was developed by Robert Shiller.  The CAPE deflates earnings and stock prices and then averages earnings over a decade, generating a P/E that is designed to capture the underlying trend in real earnings.  Note that the P/E rose from 1950 to 1965 even though rates rose.  However, as inflation steadily increased, interest rates and the P/E moved in opposite directions.  Casual observation suggests that rates above 4% and rising lead to a lower multiple.

How high will interest rates rise?  Our broad 10-year T-note model puts the fair value yield at 1.75%.

The model uses fed funds, the 15-year moving average of CPI (an inflation expectations proxy), the yen/dollar exchange rate, oil prices and German bond yields.  The most important variable keeping the fair value low are German bond yields; removing those from the model boosts the fair value yield to 2.42%.  In a less globalized world, the impact of foreign rates might be reduced, so there is a concern the model is underestimating the fair value yield.  However, as long as capital flows remain open, the impact of lower German yields should be a bullish factor for long-duration Treasuries.  In addition, if we assume a 25 bps hike in fed funds next month, the fair value yield would increase to 1.84%.

Overall, a case can be made that the recent spike in long-duration yields is overdone, at least in the short run.  On the other hand, as we discussed in the most recent WGR, if the U.S. retreats from the superpower role, inflation expectations will likely rise and weaken the case for holding long-duration instruments.  We continue to closely monitor the fixed income markets but it does appear that the long bull market in Treasuries may be coming to an end.

View the PDF