REIT Strength Means Weakness For Stocks?

I saw this chart on the Humble Student Of The Markets blog, which is one of my favorite blogs for a quick and thoughtful market perspective:

Source: 10 or more technical reasons to be cautious on stocks

The corresponding analysis is the following:

On an absolute basis, this chart of VNQ, which represents REITs, is undergoing a bullish cup and handle upside breakout. Leadership from such a defensive sector is another technical warning of a potential market top. Source: 10 or more technical reasons to be cautious on stocks

While I agree with most of the things written in the article, I have to add that the behavior of REITs at significant market tops is not as clear cut as it seems. Sometimes they outperform, sometimes they underperform the market. What is more important for REIT outperformance is the rates environment, i.e. whether rates are falling or rising. I tried to capture this with the simplest of all, the 10y US Treasury yield vs REITs:

Source: Bloomberg

First of all, during the past 10 years, we can see outperformance and underperformance as well (rising and falling white line), while the S&P 500 has been steadily rising. What is more important, is that the white line (VNQ relative to the S&P 500 index) seems to be negatively correlated with the green line (10y US Treasury), aka they move inversely. The latest period of outperformance (circled) has been coincident with the rise of the stock market from its lows in March, however if you look at the green line, it seems that it has been likely driven by the consolidation of the 10y Treasury yield. Since REITs are more rate-sensitive, this might have warranted some sort of outperformance, however its extent is quite muted.

It is not to say that the outperformance of defensive sectors is something not to worry about, but in the case of REITs there might be other factors in the background, other that a recession on the horizon. But in the past 10 years, corrections in the market have not always been preceded by the outperformance of REITs.

 

Disclosure: We are long VNQ.

2017 Review

Having finished compliance stuff for MiFidII and PRIIPs, let’s take a review of the year that passed.

I couldn’t have written a better memo myself than the guys at Hoya Capital: REITs End 2017 On High Note

What’s more, they have also collected the 2017 returns for the most important US real estate sectors and ETFs:

Source: REITs End 2017 On High Note 

So I’m just going to present some charts here that are worth watching in 2018…

  • VNQ, alias Vanguard REIT ETF which covers the US REIT sector without cell towers was rangebound all year swinging between 80 and 86 and closing almost where it started the year:

Source: Bloomberg

I’ve plotted two indicators on the chart as a textbook example of the usage of the oscillators and the Bollinger Bands. Oscillators (here the Stochastic) can best be used in trading ranges, where one sells when the indicator line closes below a standard reading (say 70 or 80), and buys when crosses above a certain level (30 or 20). So when the oscillator has a high reading one should look for the exit, and when it is low one should look for an entry. In trending markets the reverse is usually true: high readings (or closing above a certain level) are signs of strength, while low readings are signs of weakness.

The same is true for Bollinger Bands: in rangebound markets, one sells when price closes above the upper band, and buys when price closes below the lower band, basically a mean-reverting strategy. In trending markets the reverse is true: in a strong trend price usually “hugs” the lower/upper band, and travels in the same direction for an extended period.

In this case, out of several technical trading strategies, the one using Bollinger Bands generated the highest total return (almost 43%), while buy-and-hold was essentially 0. The popular MACD and various moving average strategies (50 and 20 days) generated negative returns.

This is from Bloomberg’s backtesting functionality for VNQ and 2017:

Source: Bloomberg

The lower panel shows the cumulative PnLs for different strategies, where the one using Bollinger Bands stands out as a clear outperformer.

And this is how the various real estate sector ETFs’ 2017 performance looks like:

Source: Bloomberg

The performance differences (which are huge) can be traced back to the inclusion of the cell tower segment (see this post for further details).

  • XHB (Homebuilders ETF) and ITB (Home Construction ETF) were the stars of the year, easily leaving behind every other vehicle in 2017. They have correlated better with the rising large-cap equities, while REITs acted like a bond proxy.

Source: Bloomberg

Source: Bloomberg

What a year this was for this sector! I’ve added the same two indicators to stress the point of using the proper indicator for a given market environment. For these trending ETFs, one would have sold early and never would have gotten back in time to profit from this trend. In other words, the previous strategies (oscillators and Bollinger Bands) were not successful in this trending environment. So what was successful? The easiest strategy:

Source: Bloomberg

In the lower panel of the above chart, you can see that those strategies that worked for VNQ didn’t work here, and those that generated losses there made outsized gains here.

The upshot is that you should apply your trading/investing strategy to the appropriate market environment.

  • Finally, let’s look at the performance of the mortgage-backed bonds market represented by REM, iShares Mortgage Real Estate Capped ETF:

Source: Bloomberg

This chart shows us a trending move until mid-year, when it started a lengthy consolidation. Being composed of mortgage-backed bonds, it is more or less the mirror image of the 10 year Treasury bond’s yield:

Source: Bloomberg

This is for the US market regarding the past, I’ll come back shortly with a review of international real estate.

 

Disclosure: We are long VNQ, IYR, XHB and ITB.

Performance Difference Drivers For Same-Sector ETFs

This article on etf.com provides a good summary of the question I’ve been looking at lately:

“The Vanguard REIT ETF (VNQ) has $34 billion in assets, significantly more than the $2.5 billion in the Real Estate Select Sector SPDR Fund (XLRE), which tracks the aforementioned S&P 500 sector index and the approximately $400 million Fidelity MSCI Real Estate Index ETF (FREL); however, VNQ’s 3.6% total return was below the gains for XLRE (8.6%) and FREL (6.8%).” Source: Vanguard REIT ETF Getting Makeover

How can these newcomers outperform the old beast? Probably because they track different indices, but they all supposed to track the Real Estate sector, right?

To understand the performance difference on must look under the hood, a.k.a. look at the holdings of each ETF. It is normally enough to check the top 10 or 20 stocks by weight to find the answer.

Source: Bloomberg

What’s striking is that VNQ has zero weight for American Tower (AMT) and Crown Castle International (CCI), two cell tower REITs. What’s more, the fund XLRE seems to be more concentrated in these names than the rest.

So our hypothesis now is that probably these two stocks have done better than the rest to cause XLRE and the two other to outperform VNQ this year. Let’s see if it is true!

These are the daily charts of the two stocks, with the lower panel depicting the relative performance of each stock to the VNQ etf:

Source: Bloomberg

Source: Bloomberg

The two have been massively outperforming the rest of the sector (represented here by the VNQ etf) since January, as can be seen by the rising red line in the lower panel. The price broke out of a post-election retracement consolidation range at the same time as the relative performance line broke above the descending trendline.

In light of this, we can easily explain why XLRE (and the two other ETFs) has been trailing VNQ since the beginning of September! Each of them has retraced some of its gains after breaking out to all time highs in August.

Source: Bloomberg

So in terms of fund selection, it comes down to whether these two will continue outperforming their peers or not. At least until the benchmark change in 2018….

Disclosure: We are long VNQ.