Posts
Lets start with a chart of the dividend yield. What I find very interesting is that we are at the level of dividend yield that preceded the financial meltdown in 2007. This chart on its own is extremely bearish from a valuation point of view but no doubt the bulls will say that we are at record low interest rates and this dividend yield is being viewed out of context.
So lets see whether the relative interest rate argument stands up to scrutiny. In the plot below we take a look at the spread between the REIT dividend yield and the 10yr Treasury Bond yield. The red line is the median with the blue lines representing 1 std deviations from the median. what is apparent from this metric is that although we are on the slightly expensive side, we are by no means at any valuation extreme. (in case you were wondering the pink line represents 2 std deviations).
Now there is an argument to be said that shorter term interest rates are at historic lows and this is creating the liquidity in the sector, or maybe better expressed the positive spread. With risk hurdle rates so low the incentive to invest in assets with a reasonable yield is extremely attractive, assuming the risk is being adequately priced into the asset. In the chart below we look at the spread between the REIT dividend yield and the 3month government bond yield. Here again the valuation is only marginally on the expensive side.
Something investors need to take into consideration is that the short term spread has very little room to move with the yield almost 0%. The longer term 10yr bonds are also near historic lows. Against this backdrop surprises are to be on the upside (if that is a surprise), meaning we are likely to see higher interest rates across the yield curve, which assuming dividend yields stay constant will see a compression in spreads, driving valuations to extremely overbought.
OK, we now have a general valuation landscape in place but what about the momentum we have enjoyed since the market bottomed in March 2009. In the chart below we have the total return (price+dividend) comprising SNL US REIT index. The red line is the 50 day exponential moving average(EMA) and the blue line is the slower 200d EMA. A good guide to longer term momentum is when the 50d EMA > 200d EMA it is bullish, which is currently the case, and bearish when the moving averages are in reverse order. A potential note of interest is the Relative Strength Index diverged when the index made its recent high. This could reflect a waning in the upside momentum with a potential reversal around the corner ((a clue not a sign); extra brownie points for those interested in the collective unconscious see Carl Jung's "Collected Works" to understand the difference between a symbol and a sign).
The last chart in our bag of tricks to consider today is the mean reversion characteristics in the medium term. I like to use Bollinger Bands set at 2std deviations of the 50 day look back window to see if there is a breakthrough. As you can see these bands have been able to contain at least 98% of the movements even through the global financial crisis. From the chart in its present form the market is certainly not pushing any momentum extremes.
Summary
I think it would be safe to say that the interest rate environment holds the clue to how the US REIT market will unfold from here. We are a week away from the end of QE2, how will the bond market react to the biggest buyer(s) in the market leaving. In this instance I am ranking Bill Gross's PIMCO behind Ben Bernanke's US FED. If one goes through John Hussman's brilliant analysis, the math points to the definite increase in rates in order for the FED to reduce its balance sheet and base money in circulation. There are very few who can explain things like Hussman.
Even a "moderate" program of $200 billion more Fed purchases would simply create even greater extremes, and would compound the already enormous difficulty that the Fed will face at the point that it attempts to normalize Fed policy in the least. Below, I've updated the liquidity preference schedule, reflecting nearly 70 years of monthly U.S. data. Notice that the willingness to hold base money is tightly related to the level of interest rates. More importantly, notice the non-linearity of the relationship. In order to achieve a non-inflationary increase in short-term interest rates to even 0.25%, the Fed would effectively have to reverse the entire amount of purchases it executed as part of QE2. Extending this policy to QE3 would only amplify the disruptive effect of any future Fed unwinding.
(This chart is the property of Hussman Funds)
Geek Note - R code:
The charts in this post were generated using the following code, this is my latest passion but I am a beginner so this may be very primitive, also I haven't included the manipulation of the data into zoo object:
library(zoo)
library(xts)
library(quantmod)
library(PerformanceAnalytics)
# Analysis of the dividend chart.
m<- median(dividend)
sd< sd(dividend)
chartSeries(dividend)
abline(h= m, col='red')
abline(h= m+sd, col= 'blue')
abline(h= m-sd, col= 'blue')
abline(h= m+sd*2, col= 'pink')
abline(h= m-sd*2, col= 'pink')
# Analysis of the spread of the dividend and short end of the bond yield curve.
m<- median(spread.SNLind_3m)
sd<- sd(spread.SNLind_3m)
chartSeries(spread.SNLind_3m)
abline(h= m, col='red')
abline(h= m+sd, col= 'blue')
abline(h= m-sd, col= 'blue')
abline(h= m+sd*2, col= 'pink')
abline(h= m-sd*2, col= 'pink')
chartSeries(SNLindex.price)
# Analysis of the spread of the dividend and long end of the bond yield curve.
m<- median(spread.SNLind_10y)
sd<- sd(spread.SNLind_10y)
chartSeries(spread.SNLind_10y)
abline(h= m, col='red')
abline(h= m+sd, col= 'blue')
abline(h= m-sd, col= 'blue')
abline(h= m+sd*2, col= 'pink')
abline(h= m-sd*2, col= 'pink')
# Analysis of the shape of the bond yield curve.
m<- median(spread.10y_3m)
sd<- sd(spread.10y_3m)
chartSeries(spread.10y_3m)
abline(h= m, col='red')
abline(h= m+sd, col= 'blue')
abline(h= m-sd, col= 'blue')
abline(h= m+sd*2, col= 'pink')
abline(h= m-sd*2, col= 'pink')
# Analysis of Mean Reversion
chartSeries(SNLindex.price)
addBBands(n=50, sd =2, ma = "EMA")
chartSeries(SNLindex.price)
addMACD()
chartSeries(dividend)
addBBands(n=50, sd =2, ma = "EMA")
# Analysis of Different Trending Patterns
chartSeries(SNLindex.price)
addEMA(n=200, wilder= TRUE, col= 'blue')
addEMA(n=50, wilder= TRUE, col= 'red')
addRSI(n=16, wilder= TRUE)
