On Positive Bond-Stock Correlations, And Ricardo's Nightmare
According to Goldman's Michael Vaknin, the positive correlation between bonds and stocks is somewhat odd, and he explains it by claiming that stocks are going up on expectations of economic growth as bonds drop in advance of the Fed's "trivial" elimination of bond supply (we will see how trivial it is when there are no Treasurys left to buy in one year). Well, he is half right: the only trade remaining is what we have been claiming since the beginning of the year - frontrunning the Fed. Which explains why the biggest lament at the Value Investor Congress ealier this week (where Maverick's Ainslie was praising Apollo Group a little prematurely) was that there is, well, no value investor left - all that is left is parsing the H.4.1, the H.3 and the H.6 Fed statements. Everything else is irrelevant. Vaknin does, however, bring up an interesting point, one which has made both Bank of America and JPM highlight the bubble in Emerging Markets: namely that the Fed is exporting low yields abroad. Which also is logical: emerging countries are hoping to offset productivity loss due to reduced exports via appreciating market values. However, as the latter is merely an artifact of Keynesian FX warfare, it simply can not last, as at the end of the day everyone focuses on transitory relative strength instead of doing what Ricardo so long ago correct predicted: each country must emphasize what it has a competitive advantage in. It appears lately the only competitive advantage that is important is who can print the most fiat the fastest. And that is a recipe for a complete wipe out.
Full thoughts from Goldman's Vaknin:
The stock shall live with the bond….at least at this stage of the cycle
One of the interesting features of the price action in financial markets in recent months has been the positive correlation between stocks and bonds. Indeed stocks have started recovering in early 2009 already, and while bond yields have experienced local lows at the time, the price action of the recent months has now brought benchmark yields to new cycle lows.
The current environment is not all that different form cyclical behaviour of stocks and bonds historically. Indeed, last year Dominic Wilson and Kamakshya Trivedi showed that stocks and bonds have tended to positively co-move at the stage in the cycle, where growth rebounds, yet inflation remains low or declines further (as with the current backdrop). In this environment bonds remain broadly supported by a very benign inflation dynamic (which typically tends to turn around only a year or two after the pick up in growth). Stocks, in addition to being supported by a very low discount factor, would typically move to reflect the prospects of growth acceleration translating into strong earnings.
The exception to this historical 'norm' was the dot.com cycle, when stocks turned up in 2003:Q1 - 'only' one quarter before rates bottomed out. The reason was simple – inflation re-accelerated faster than has been the case in previous cycles. Indeed back then core inflation moved from the 1.1% trough level to above 2% in less than one year.
This time around, with spare capacity in labour and housing markets much higher, the Federal Reserves sees the current inflation backdrop as dangerously low. This backdrop is precisely why the Fed has been signalling that a second round of quantitative easing is underway. Such a policy action has also been supportive of the positive correlation between bonds and stocks - stocks rally on the implicit message that policy rates will remain exceptionally low for longer, while bonds are trivially supported by the fact that the Fed will reduce available supply in the market.
The export of low yields – another potential source of positive bond-stock correlation
What seems to be also relevant for the bond-stock correlation this cycle is the notion that the rally in benchmark bonds has been 'exported' to EM and G10-peripheral economies in a meaningful way.
Indeed, it is evident that the real policy rate in EM and G-10 peripheral countries are now significantly lower than the trough seen in the previous cycle, unlike the 'core' countries for which the current real rates environment is broadly similar. Indeed based on 1-yr ahead real rates (using 1-yr swaps minus 1-yr consensus inflation expectations) we note that EM and G-10 peripheral rates are now remarkably low at -0.5% and -0.4% respectively. By a comparison, the troughs in real rates in the dot.com cycle were considerably higher at 3.0% and 1.0% respectively. In the US and in Euroland on the other hand, real rates are now at -1.2% and -1.3% respectively - just a touch below the -1.0% cycle trough (for both) back in 2002.
So clearly the anomaly here is the significant extent of policy accommodation outside the 'core' countries. While in theory 'non-core' central banks should have tightened monetary policy more aggressively, rates remained relatively low, partly in order to avert substantial FX appreciation (this point was discussed extensively by Dominic Wilson and Stacy Carlson in our latest Global Economics Weekly). This is probably why 'non-core' stocks have generally outperformed those of the core economies, and in many instances, they seemed to be leading the rally in the 'core' stock markets.
One interesting way of gauging whether the outperformance of 'non-core' equity markets is indeed linked to exceptionally easy monetary policy is to look at the composition of growth in these economies. If indeed growth is supported by low real rates, one should expect to see interest rate sensitive sectors leading the way.
Indeed a close inspection of the recent growth dynamics is very much consistent with this hypothesis. Relative to the 1997-2007 average, domestic demand growth in the G10 periphery is now 0.5% higher than the long term average of about 2.5% per year. Likewise, EM domestic demand growth is now 1.5% higher than the average (6.5%). By comparison, domestic demand growth in the US is 0.6% lower than the long term average (3.4%). More importantly, next year, our US economists are forecasting a considerable deceleration in domestic demand to 1.6% growth – around half of the average growth rate.
The picture looks broadly similar when we drill down into the consumption component. Indeed relative to the average, consumption growth is running 0.2% higher in the G-10 periphery and 0.6% higher in EM. By contrast, US consumption growth is below the trend by a significant 1.8%. Focusing on the investment portion of domestic demand, we note that investment growth remains very week in the US - annualised growth is running 1.4% below the trend. In the G10 periphery investment is comparatively better at 0.9% below the average, while in EM, investment is much more impressive – running 1.4% above the trend.
All told, in addition to the “normal” positive relationship between bonds and equities at this point of the cycle, the positive correlation is likely getting a boost from strong growth in EM and G10 peripheral countries, which keep their policy settings unusually accommodative.