Friday 25 July 2014

Low Volatility ETF Update - Are they performing as advertized?

In 2012 we introduced the newly launched category of low volatility equity ETFs that promised more stability and equal or better returns in comparison to standard index funds with holdings selected and weighted by market capitalization. A couple of years have passed, so let's see how they are performing.

Comparison Table - Low Vol ETFs vs Cap-Weight Benchmarks
(click to enlarge table image)


Low Vol performance mostly lags Cap-Weight ...
The green cells in our comparison table above show where the Cap-Weight ETFs have done better than their Low Vol counterparts. The green cells outnumber the yellow cells showing Low Vol superiority. It is interesting how consistent is the pattern of returns. In every single equity market around the world, Cap-Weight has higher returns for the year ending June 30th. But the tide seems to have turned since the beginning of 2014 - most of the Low Vol ETFs have pulled ahead Year-To-Date (YTD).

... and the reason is that Cap-Weight should win in a strong bull market
As we noted in the original post (and as reiterated in this excellent National Bank Financial January 2013 analysis of Low Vol ETFs), the main drawback of Low Vol strategies is likely performance lag in strong upward markets. That is what has been happening - returns in every world market have been wonderful recently - 14% to 28% 1-year total returns in the various global equity markets. The Low Vol returns have been very good too, just not quite as good.

More evidence of how good things have been recently shows in the columns for Reward to Volatility (Sharpe) Ratio and Volatility (standard deviation of returns). Sharpe ratios are very high - compare the 1-year numbers ranging from 0.6 up to 4.6 versus the 10-year averages of 0.4 and 0.5. Historical long term Sharpe ratio figures across many equity markets cited in the 2012 Credit Suisse Global Investment Returns Yearbook (see Fig.4 on page 19) range from 0.1 to 0.7. We have been living in unusually good and stable times for the last few years. Volatility itself is also very low compared to past longer term averages. Equity volatility below 10 is well below the usual high-teen or greater long term average (see Tables 1 and 2 on page 22 of Credit Suisse). The result is that in several cases, the Cap-Weight ETF benchmarks actually display lower volatility than the Low Vol ETFs which are specifically designed to be lower volatility.

When markets go sour again, that will be the real test of Low Vol ETFs
It is perhaps obvious to say it but the current equity party will end - market volatility will return and large market drops will happen. There's no reason to think the history of markets will not be repeated and equities will not exhibit considerable volatility in future. The only thing we don't know is exactly when that will happen. That is when we will see how much the Low Vol ETFs limit the downside drop and the volatility, perhaps proving their worth with better risk-adjusted (higher Sharpe ratio) and total returns.

We investors must decide whether to wait another five or ten years for definitive proof to emerge to substantiate the benefits of actual ETFs that attempt to capture results shown in academic research and backtests. Waiting means possibly missing out on Low Vol benefits ... or missing out on Low Vol disappointment if they really do not pan out.

As far as what other investors are doing, so far, the net asset figures of the various ETFs in our table show that Low Vol has caught on mainly in the USA. In Canada, only BMO's low vol version has attracted substantial investor money.

Disclosure: This blogger owns shares of the BMO Low Volatility Canadian Equity ETF (TSX symbol: ZLB).

Disclaimer: This post is my opinion only and should not be construed as investment advice. Readers should be aware that the above comparisons are not an investment recommendation. They rest on other sources, whose accuracy is not guaranteed and the article may not interpret such results correctly. Do your homework before making any decisions and consider consulting a professional advisor.

Monday 21 July 2014

Inflation Protection's Big Trade-off - Protection vs Returns

Friday the latest CPI inflation indicator for Canada showed a 2.4% rise. Inflation has been fairly low and quite stable for more than a decade now but the wary investor keeps the more distant past in mind when inflation ran amok and therefore factors an amount of "just-in-case" protection against any resurgence and surprise spikes of inflation.

As with many situations there is a trade-off - what works very well as protection often has low returns. Let's therefore have a look at the major asset classes to see how they respond to unexpected inflation.

Expected vs Unexpected Inflation - First, it's necessary to clarify that what's expected won't hurt so much, it's the unexpected that matters. With inflation running at around 2%, the expected returns and yields of various investments have that amount (as we noted in this recent post) incorporated into prices. It's the shock and surprise increases that take away real value which hurt investors.

Macro-Economic Volatility and Supply Shock Inflation vs "Great Moderation" Inflation - It probably has not seemed so to those suffering through rising prices, but the nature of inflation has differed in the 1990s and 2000s from the decades of the 1970s and 80s. The less volatile business cycles of more recent times, termed the Great Moderation, have entailed a different kind of inflation. In the 1970s, inflation arose more from supply shocks, in the form of big oil price increases, and at times when the economy struggled, while in the more recent period inflation has risen in tandem with good economic growth. That, according to the research paper Inflation-Hedging Portfolios: Economic Regimes Matter (on SSRN) by Marie Grière and Ombretta Signori, has meant that the best protection against inflation has changed, as we note below.

Short-term vs Long Term - The final key distinction is that the type of investment that responds well and quickly to a surprise inflation leap may not do as well as another over many years.

Asset Classes - How they respond to inflation

a) Cash - Except for a short few months of typical lag before rates adjust, cash investments like T-Bills and money-market funds will rise with inflation, no matter what the economic regime. The bad news is the trade-off that returns on cash short- and long-term are low and, as Brière and Signori point out, the return on a long-term investment in cash stands a fair risk of falling shy of full inflation-matching. BlackRock's review Are Hedging Properties Inflated? notes how central bank policies like ones in place at present may mean negative overall real returns for cash.

b) Nominal Bonds (Government 10+ year) - Bonds do not usually offer inflation protection. On the contrary, in the short term bonds suffer losses through price reductions. However, with longer holding periods extending out more and more years, bonds progressively recover. During the years of the Great Moderation, Grière and Sognori even found nominal bonds to be a good long-run inflation hedge, whereas in the earlier opposite economic environment, nominal bonds were a bad choice for inflation compensation. This brings to the fore a key issue for investors - will the Bank of Canada (and other central banks) continue to maintain their credibility and the belief of investors that inflation will not be allowed to run rampant? If inflation is consistently contained, nominal bonds will do fine as inflation protection. The advantage of nominal bonds over inflation-indexed bonds of similar duration is slightly higher yields.

c) Inflation-Indexed (Real Return) Bonds - Such bonds, called Real Return Bonds in Canada and TIPS in the USA, offer explicit inflation hedging by the built-in mechanism to increase interest and principal in lockstep with a measure of inflation, in Canada that being the CPI. Any and every inflation rise, whether expected or unexpected, feeds automatically into the bonds' returns three months after the actual CPI results. The downside is that net real returns after taking away the inflation-matching portion are very low.

d) Equities (Broad market cap-weight index) - Equities are hurt by unexpected inflation in the short term and especially so when inflation is of the supply shock variety, as the 1970s oil price shocks so well demonstrated when stocks had negative total returns while CPI rose significantly (e.g. see this chart of the USA's S&P 500 index). In the long term, the ability of businesses to cope with inflation by passing through price increases to make up for inflation has enabled gradual, though sometimes not full recovery depending on the period as BlackRock show. The Credit Suisse Global Investment Returns Yearbook for 2012, which examined the inflation-fighting performance of various asset classes, summarized equities as "not particularly good inflation hedges". On the other hand, equities offer the best long-term historical returns record.

e) Real Estate - This asset class is a very poor short-term inflation hedge but an effective long term hedge (see all our reference studies linked in this post), with returns similar to equities.

f) Gold and Precious Metals - In the short-term, gold has been found to be a very good hedge for unexpected inflation but only erratically in certain periods, and is a weak hedge for overall inflation in the long term. It is an asset that produces no income and its long term return from capital appreciation only is very low as the Credit Suisse Yearbook data shows. It's hedging value is probably more related to financial and currency collapse than inflation.

g) Commodities - There appears to be consensus that commodities offer excellent short-run inflation hedging according to an IMF paper of 2011 Inflation Hedging for Long Term Investors by Alexander Attié and Shaun Roache. BlackRock ranks commodities as the best short-term inflation hedge of all the asset types. This conclusion applies to broad commodity indices as individual commodities are not effective. The protection offered by commodities erodes over time till at more than five years, there is no benefit. Worst for the long term, commodities have exhibited negative real returns far short of inflation as the graph below copied from the Attié and Roache paper shows where the dark black line of commodities lies below CPI inflation.
(click to enlarge image)

Take-aways
Given 1) the variable hedging performance of the various asset classes, 2) the difficulties in knowing when inflation surprises are set to arrive (no kidding!) and what type of economic regime and inflation drivers there will be and 3) the need to gain higher returns than simply keep up with inflation, the sensible thing to do for the investor with a long time horizon would seem to us to include some of each asset, to maintain a diversified portfolio that will cope with all the eventualities. Rebalancing to the target asset allocation will take into account the effects of inflation reactions as well as the many other effects affecting asset values such as real interest rates, changing perceptions of riskiness and foreign currency swings.

Disclaimer: This post is my opinion only and should not be construed as investment advice. Readers should be aware that the above comparisons are not an investment recommendation. They rest on other sources, whose accuracy is not guaranteed and the article may not interpret such results correctly. Do your homework before making any decisions and consider consulting a professional advisor.

Saturday 12 July 2014

Dogs of the Dow and the TSX - Are they nasty or nice investing pets?

It's now the dog days of summer, so it's a good time to write about an investing theory referring to dogs that seems to have worked pretty well since it was first put forward by Michael O'Higgins in his 1991 book Beating the Dow.

The theory is called the Dogs of the Dow. The method is ultra-simple: once a year, take the list of the 30 stocks in the Dow Jones Industrial Average, sort them by dividend yield, and buy an equal amount of the ten highest yielding stocks. Hold them for a year and then the following year, sell the stocks no longer in the top ten and buy the replacements. By convention the buying and selling is done on January 1st, though there is nothing to prevent anyone from starting at any time during the year.

The popularity of the strategy has even led to the creation of a dedicated Dogs of the Dow website, where the 2014 January 1st list of Dogs is shown in blue. For someone buying today, a stock screener such as Finviz.com with the DJIA index selected, will list all its components, which can then be sorted by yield.

The logic and rationale of the Dogs method is that the dividend yield indicates stocks which are out of favour, usually only temporarily. Dividend yield is defined as dividends over current stock price so the yield of a stock rises when price declines i.e. the stock falls out of favour. Thus, the Dogs strategy is often described as a simple, crude strategy for identifying Value stocks.

Performance - Up-to-date total return data seems difficult to find, but up to 2011 the Dogs strategy performed the same as the overall DJIA with a 10.8% average over the preceding 20 years according to Dogs of the Dow. Since then, the Dogs have edged ahead, judging by the price in blue in the chart below of the sole ETN that tracks the Dogs, the Deutsche Bank AG ELEMENTS Dogs of the Dow Total Return Index Note (NYSE: DOD, MER 0.75%) compared to the red line of the DJIA.
(click image to enlarge)



Criticisms and caveats:
  • Equal weighting of selected Dogs while the DJIA is price-weighted gives the Dogs an unfair advantage (see Wikipedia write-up and Forbes article) i.e. the performance comes from the weighting not the stock selection
  • Dividends alone are no longer a good test of fundamental value (e.g. this article in Barron's magazine) due to a rising rate of share buybacks; instead something like shareholder yield (net of share buybacks minus issuance plus dividends) would be more appropriate
  • Lack of diversification of the Dogs as a portfolio due to the small number of stocks and accidental sector concentrations
Dogs of the TSX
The same idea has been applied in Canada on stocks in the TSX 60 Index, though with some modifications by the originator of the Canadian version Dave Stanley, like the exclusion of deliberately high-yielding income trusts and former trusts (as described in the Beating the TSX Finiki article and as discussed in this thread on the Financial Wisdom Forum), to ensure a similar focus on blue chip, large cap stocks.

Up to 2011, the Dogs of the TSX did even better than the Dow original, beating the TSX 60 Index by more than 2% per year on average between 1987 and 2011 (according to Finiki, citing a 2012 Canadian MoneySaver article).

Current List of TSX Dogs:
Using the superb Stock Rover screener, we extracted all the high yielding larger cap stocks on the TSX, then filtered out all stocks that are not included (rows highlighted in pale yellow) in the TSX 60 Index, as well as any income trusts and former income trusts (rows in tan) not already excluded to come up with the table below. The white rows are the TSX Dogs as of July 11.
(click to enlarge)

The rules of selecting the Dogs are not set in stone. It's an active strategy, not a passive one and we investors can change what we want.  Some of the companies may be in the doghouse for good reason. Norm Rothery of Stingy Investor has just published in MoneySense his list of Top Dogs of the TSX, filtering out stocks where the dividend may be in danger, as indicated by dividends that exceed earnings. Notice how TransAlta (TSX: TA), top of the list above with highest yield, got dropped, no surprise given its negative earnings. Some investors might consider TransAlta a good buy as a turnaround candidate but others, who want only to buy high quality blue chip companies and their dividends, may well want to exclude TA, which is not even a very large cap company.

Other differences between our list above and Rothery's probably arise because his list came from data as of June 30th. Two on his list - Telus (TSX: T) and Power Corporation (TSX: POW) - are 11th and 12th in line in the Stock Rover screen. Market prices are always changing.

The resulting Dogs looks less like ones that will bite you and more like faithful companions. Most of the current Dogs have shown up in our previous posts of centenarian companies (National Bank, CIBC, BCE and Bank of Montreal and blue chips (as well as the centenarians, Potash Corp, Rogers, Shaw) or dependable utilities (Fortis).

Other variations of the Dogs selection method are also possible e.g.:
  • Relax the exclusion of (former) income trusts, especially now that it has been several years since the 2011 trust conversion deadline and excessive payouts would most likely have had to be reigned in. For example, Canada's largest REIT Riocan (TSX: REI.UN) has a high yield which is still well under its earnings, as our table above shows. Global Securities writes about such an apparently successful implementation of this Dogs variation.
  • Allow in large caps that the TSX 60 excludes, like Power Financial (TSX: PWF), Great West Life (TSX: GWO) and IGM Financial (TSX: IGM)
Due to the lack of diversification / the high sector concentration (4 out of 10, or 40% on an equal investment basis, are in Communication Services, with none in Industrials or Consumer stocks and only 10% each in Energy and Materials), we would not think it advisable to put all a Canadian equity allocation into this strategy. However it does offer a way to identify stocks that currently offer higher dividends at a reasonable price and to establish long term positions in high quality companies.

Disclaimer: This post is my opinion only and should not be construed as investment advice. Readers should be aware that the above comparisons are not an investment recommendation. They rest on other sources, whose accuracy is not guaranteed and the article may not interpret such results correctly. Do your homework before making any decisions and consider consulting a professional advisor.

Tuesday 8 July 2014

Long Term Return Expectations for USA and Rest of World - the Current Outlook

Last week we laid out what by historical standards are the modest future long term returns investors can expect from Canadian stocks and bonds. Since most Canadian investors also include an allocation to equities of the USA and other countries, this week we'll review the indicators of their likely future returns over the next 10+ years.

US Equities

Method 1 - Current Dividend Yield + Dividend Growth
The idea is the same as that in our post about Canadian stocks but since US companies make heavy use of stock buybacks (see this Credit Suisse article) to return money to shareholders instead of dividends, we'll adopt the modification that takes account of buybacks and new share issuance to derive Shareholder Yield. Patrick O'Shaughnessy at Millenial Invest recently explained Shareholder Yield and graphed its historical actual values for the US equity market, which currently looks to be just under 4% - we'll use 3.9%.

Then we add potential dividend growth based on growth of the overall US economy. Forecasts of long term US GDP growth vary from the Conference Board's Global Outlook of about 2% as an average of the two periods from 2014-19 and 2020-25, to the US Federal Reserve's 2.2 to 2.4% cited by Rick Ferri back in January. We multiply those GDP forecasts by the 0.6 proportion that flows through on average to shareholders (as we reported in our original post on future return expectations, research shows that GDP growth gets diluted and does not translate into as much shareholder return) to get a range of 1.2% to 1.4% annual dividend growth.

Result: A reasonably attractive, at least compared to what CAPE tells us below, total US real annual equity return estimate of 3.9 + 1.2/1.4 = 5.1 to 5.3%

Method 2 - Shiller's Cyclically-Adjusted Price Earnings (CAPE) Ratio
Yale professor Robert Shiller found that a P/E that adjusts earnings for inflation and averages over the past ten years smooths out economic cycles in calculating a Price/Earnings ratio for the US stock market. The adjusted P/E or CAPE gives a good indicator of probable future long term stock returns. The higher the CAPE the lower the future returns. Clifford Asness of AQR Capital explains the ratio in more detail and defends its usefulness in An Old Friend: The Stock Market’s Shiller P/E.

The CAPE is now over 26, which is very high by historical standards, as this chart from Guru Focus shows. Note the tiny implied future annual return, a mere +0.6%!
(click to enlarge image)



The following chart from a post at the Blakeley Group shows what happened to subsequent S&P 500 returns at various CAPE levels through the past 114 years.


However, the relationship between CAPE and US equity returns has a fair degree of variability - all is not deterministically and rigidly fore-ordained. Asness includes a table showing that for CAPE over 25, subsequent annualized 10-years have ranged over a worst case of -6.1% to a best of +6.3%.

Method 3 - Ratio of Total Market Cap to GDP
This is apparently Warren Buffett's method of assessing whether the US market is over- or under-valued. As calculated by Guru Focus, it currently shows the US market to be significantly over-valued. Consequently the expected future return for US equities is a paltry 1%.

We would have hoped for closer agreement between the three methods but that's the situation. US equity real returns will be muted but there seems to be a high degree of uncertainty with a plausible range from 0.6% to 5.3%.

For those who wish to look at nominal returns including US inflation, simply add the Federal Reserve Bank of Cleveland's current 10-year inflation forecast of 1.8% to the real return figures above.

Developed and Emerging Market Country Equities
CAPE works about as well for other developed country markets as for the US, but less reliably for Emerging market countries, according to this 2012 paper by Joachim Klement. The Shiller P/E for Europe has been rising but is still low in relative historical terms as seen in the charts on MrMarket.EU and in this presentation by Henderson Global Investors. The latter two papers do not translate the low CAPE numbers into projected returns but Klement does and his projections, though two years old, may still be useful, especially for emerging markets which have gone sideways since 2012. His projections are for five years only. The annual compound real return for emerging markets he projects at 5.8%. For developed markets the projection in 2012 was 5.7% annual real growth, but we have seen gains of over 23% in the last year alone so a current projection will not be as high.

Guru Focus uses a method similar to the dividends plus dividend growth formula we have used above, but it explicitly adjusts for reversion to the mean of the ratio of market cap to GDP. By this method, of the 19 major countries in the table of CAPE-based return expectations on Guru Focusevery single country, both developed and emerging market, is estimated to have future returns significantly above the USA's (see chart below). Emerging markets have very high projected returns. Incidentally, we note that the Guru Focus estimate for Canada is 3.9%, which fits within the bound of other estimates we wrote about last week.
Takeaway:
A Canadian investor should look to maintain, or establish if not already there, an allocation in the portfolio to developed country and emerging market equities. The USA offers less promise of reward than those markets.

Caveat: A Canadian investor's return on foreign assets will be affected by moves of the Canadian dollar versus the currency of the foreign country equities. Though such moves tend to even out in the very long run for investors who maintain and regularly rebalance their asset allocation, as we discuss here and here, over longer periods such as ten years or more, returns may raised or reduced by foreign exchange moves. Forecasting foreign exchange shifts is too uncertain to be worth trying.

Addendum April 2016: McKinsey & Company's Global Institute is forecasting much lower returns for the USA and Europe than those experienced over the last 30 years - 1.5 to 4.0% per year lower on equities and 3 to 5% on bonds. Research Affiliates provides monthly updated forecasts of long term future returns for many asset classes, including Canadian equities, which as of the 31 March 2016 update pegged the 10-year expected real return at 3.4%.

Disclaimer: This post is my opinion only and should not be construed as investment advice. Readers should be aware that the above comparisons are not an investment recommendation. They rest on other sources, whose accuracy is not guaranteed and the article may not interpret such results correctly. Do your homework before making any decisions and consider consulting a professional advisor.