There are few things that we can
control when it comes to investing; asset allocation is one of them,
and it may be the most important. Asset allocation is the exercise of determining
how much of each asset class you should hold in your portfolio. In general, the asset classes
that we have to choose from are stocks, bonds, real estate investment trusts, and alternatives.
Those categories can be broken down further, but I will leave it there for now.
Except in hindsight there is no optimal asset allocation. The best that we can do is
take guidance from the academic literature. I’m Ben Felix, Associate Portfolio
Manager at PWL Capital. In this episode of Common Sense Investing, I’m going
to tell you about asset allocation. Building on the work of Harry Markowitz, William Sharpe and John Lintner are credited
with developing the Capital Asset Pricing Model, or CAPM. If you have taken any finance courses
you are at least aware that the CAPM exists. This was the first model that quantified the
relationship between risk and expected returns. The CAPM is a single factor model. It sees risk
and return as being determined by a portfolio’s exposure to market beta, or the riskiness
of the market as a whole. Market beta is a priced risk. In other words, we expect a
positive outcome for maintaining exposure to market beta. This is very different from
betting on a specific segment of the market. That introduces idiosyncratic risk, which
does not have a positive expected outcome. To eliminate idiosyncratic risk we start
our asset allocation journey with the market portfolio. A total market index fund
is representative of the market portfolio. The next question is which
stock markets do we need exposure to? The answer, in short, is all of them. Combining Canadian, US, International Developed,
and Emerging Markets stocks into a portfolio improves the risk and return characteristics
relative to each of the individual parts. This would be expected considering the imperfect
correlations between each asset class. US equities are an exception as they actually
look better on their own over many historical periods. We know this in hindsight,
but should not bet on it going forward. The optimal mix between Canadian,
US, International developed, and emerging markets stocks in a portfolio
is an unknown. This makes the geographic asset allocation choice mostly arbitrary,
as long as global exposure is achieved. Most model portfolios, including PWL Capital’s,
Wealthsimple’s, Vanguard’s asset allocation ETFs, and the Canadian Couch Potato portfolios have a
heavy bias toward Canadian stocks, at about ⅓ of the equity allocation. This might be surprising
when it is considered that Canadian stocks make up about 3% of the global market cap. Why so
much in Canada? It comes down to tax treatment. I have talked in the past about unrecoverable
foreign withholding tax on foreign dividends in registered accounts, which is an issue that
you will never have with Canadian stocks. Further to that, in a taxable account
Canadian dividends receive preferential tax treatment. With asset allocation
being a somewhat arbitrary exercise, assuming some level of global diversification,
favouring a tax efficient asset class is sensible. We have established that it is sensible to
diversify globally to attain exposure to the stock market, and it might make sense
to be overweight Canadian stocks relative to the market to reduce the impact
of both foreign and domestic taxes. As a point for future discussion, the
Canadian dollar return for a portfolio equally split between Canadian, US and
International including emerging market stocks from January 1990 through July 2018
was 8.17% with a standard deviation of 11.92%. Now let's talk about bonds. Bonds
are much less risky than stocks, and they have correspondingly
lower expected returns. If you have noticed that Canadian bonds have
almost matched the return of Canadian stocks for the last 30 or so years, with less than half
the risk as measured by standard deviation, let me give you some context to understand
those numbers. This time period captures the greatest fall in interest rates in history.
Falling interest rates make bonds look amazing, so the numbers need to be
taken with a grain of salt. Canadian bonds have had a low correlation to our
Canada-heavy global equity portfolio going back to 1990, and adding them to the portfolio looks
great. A 10% allocation to Canadian bonds barely reduces the return, moving it down to 8.13%, while
dropping the standard deviation more than 1%, to 11.83%. Keep in mind that bond returns were
uncharacteristically high over that time period, so we might expect a bigger drop in returns
for adding bonds going forward. Bonds would typically be added to a portfolio to reduce its
riskiness, not to increase its expected returns. Finally, adding in a 6% allocation to
real estate investment trusts increases our historical annualized return to 8.32% while
lowering our standard deviation to 10.60%. As of now we have a portfolio consisting
of 10% Canadian bonds, 6% US REITs, 28% Canadian stocks, 28% US stocks,
20% international developed stocks, and 8% emerging markets stocks. I have some
comments on REITs which I will save for later. So far we have seen some big improvements by
adding in relatively small amounts of asset classes with imperfect correlations to stocks. I
am not going to talk about adding in alternatives like hedge funds, managed futures, preferred
shares, and high yield bonds. While they may be uncorrelated asset classes that look good in
a back test, they come with other unfavourable characteristics which I have discussed
in detail in past videos and blog posts. I think that this is about as far as most people
get. Choosing some mix between stocks and bonds, and maybe REITs, based on their ability,
willingness, and need to take risk. Stopping here ignores the most up to date research
on financial markets and portfolio management. In their 1992 paper the Cross Section of Expected
Stock Returns, Eugene Fama and Kenneth French summarized the body of research showing that
the CAPM has substantial shortcomings. They essentially concluded that the CAPM only explains
about two thirds of the return differences between diversified portfolios. So two portfolios with a
beta of 1 might have had substantially different returns, with no way to explain the difference
other than attributing it to active management. The following year, Fama and French
proposed a new asset pricing model called the Fama-French Three-Factor model. Instead
of relating expected returns to market risk, the three-factor model relates expected
returns to exposure to the market, exposure to small stocks, and
exposure to value stocks. This model explains about 90% of the difference
in returns between diversified portfolios. This is important for investors because if
there are three independent risk factors that explain returns, we want exposure to
all three, not just to one. As we have seen, adding in imperfectly correlated risks
should increase our expected returns and decrease our risk. More recent research
has identified at least one other factor that can be sensibly added to portfolio
construction. That factor is profitability. These four factors, market beta, size,
value, and profitability have had low and in some cases negative correlations with
each other over time. The correlations of factors with each other are even lower than
the correlations between geographic regions for stocks, and in some cases lower than
the correlations between stocks and bonds. I will quote a 2012 article published in
the Journal of Portfolio management titled the Death of Diversification
has Been Greatly Exaggerated: “ The argument that we make for factor
diversification partly rests on the expectation that the positive factor
premia will continue to persist. But the correlations (or lack thereof) these premia
with each other are at least as important”. One of the confusing things about getting factor
exposure is that holding small cap stocks in a total market index fund is not sufficient. If you
have the same amount of small cap or value stocks as the market, you only have exposure to market
beta. It is only by increasing the exposure to small cap and value stocks beyond market cap
weights that factor exposure can be obtained. Coming back to our hypothetical
portfolio, if we split up each geographic region into one third market,
one third value, and one third small, we end up with an annualized return from January
1990 through July 2018 of 9.17% with a standard deviation of 10.36%. Clearly adding in factor
exposure was beneficial over the time period. This not just me cherry picking a data to
make a point either. I can’t name all of the studies that have demonstrated the benefits
of factor exposure, but here’s one: In a 2017 study published in the Journal of Portfolio
Management, Louis Scott and Stefano Cavaglia examined the impact of factor diversification on
the odds of retirees outliving their portfolios. They found that terminal outcomes can be
significantly enhanced through factor exposure. The catch is that getting factor exposure is
not always easy. It’s great if you can get it, but it is a challenge, especially for DIY
investors in Canada, due to a lack of products. I said I would come back to REITs. Recent
research has demonstrated that the return of REITs is explained by the market beta,
size, and value factors, in addition to the term and credit factors which are factors that
explain fixed income returns. Based on that, a portfolio with exposure to the aforementioned
factors may not need an allocation to REITs. At this point I think that total stock
market exposure is a given in the asset allocation decision. The mix between stocks
and bonds is more subjective based on personal circumstances and preferences. Exposure to
factors is important but often overlooked, with the asterisk that even if you
want it, it may be hard to implement. Tell me about your asset
allocation in the comments. Thanks for watching. My name is Ben Felix
of PWL Capital and this is Common Sense Investing. I will be talking about a new
common sense investing topic every two weeks, so subscribe and click the bell for updates.
