The value of portfolio construction in fixed income

Kamyar Hazaveh's picture

Reading and writing about 2015’s outlook is a ritual for most market participants. Most banks and buy side firms publish their upcoming year’s outlook in December with forecasts on bond yields, equity returns, and a general macro outlook. Some publish the list of their top 10 or 20 risks for next year. In 2014, the major calls at the beginning of the year were for higher bond yields and acceleration in global growth. The risks to the 2014 outlook were mostly related to the scenario in which bond yields rose too much that led to catastrophic losses in bond portfolios with ripple effects across other asset classes (basically taper tantrum 2.0).

What actually transpired in 2014 was quite different. Global growth slowed substantially with deflationary headwinds. Bonds staged one of the best rallies over the last couple years with our major Canadian and global bond portfolios posting about 10% in absolute returns. The risk events that transpired in 2014 included the situation in Ukraine and the substantial decline in commodities. These were on no one’s radar at the beginning of 2014.

The future is unknowable

Over the years, I have come to realize that our industry as a whole has a spectacularly poor track record in predicting the future. The simple truth is that the future is unknowable by nature and definition. Asset class returns (e.g. cash, bonds, stocks, etc.) are unknown ex-ante and only look easy and obvious ex-post with 20/20 hindsight. The financial media, however, is full of pundits who confidently express views about the future returns of stocks, the direction of bond yields, etc.

The reason why pundits on different media outlets get air time is that it is human nature to crave certainty in an uncertain world. For almost our entire time on the planet, we have been at the mercy of the elements on earth and natural disasters. We are hard wired to seek certainty for our survival. This is behind man’s fascination with superstition and the supernatural.

People find comfort in a pundit who confidently talks about upcoming employment numbers or the next move by the Federal Reserve. In fact, studies have shown that the more confident the talking heads sound, the more coverage and followers they have. The same studies have looked into the accuracy of forecasters and expert predictions. The truth is that the track record of forecasters and pundits is very poor.

Investing in an uncertain world

Modern portfolio theory (e.g. Mean-Variance Optimization or Markowitz Portfolio) suggests that there are three inputs to building an efficient portfolio: expected returns for each asset class involved, expected volatility, and forecasts of correlations (the way different asset classes move relative to each other). In the 1970s, this was a huge step forward in reducing portfolio volatility, increasing portfolio Sharpe ratio (risk-adjusted return), and protecting investors against the risk of ruin by exploiting the diversification benefits of bonds and stocks.

This simple construction has been behind most stock/bond portfolios which usually lead to a 60/40 mix of stocks and bonds. Over the years, however, large portfolio managers have discovered the short-comings of modern portfolio theory and have improved upon it.

The major problem with Markowitz’s portfolio was that the forecasts of returns (or expected returns) proved to be unreliable and often wrong – yes, the future turned out to be unknowable. The estimation of correlations also tended to be wrong in market extremes such as the 2008 crisis.

There are ways to improve upon the Markowitz portfolio. One must admit that the most uncertain input to portfolio construction process is the forecast of returns. The first step is admitting that the future is unknowable does not mean that one should not invest. To the contrary, admitting the uncertainty about future returns is the first building block in creating an efficient fixed-income or multi-asset portfolio. The second step is understanding the response of different asset classes to major macro drivers (growth and inflation) and understanding the correlation of different asset classes in each macro-economic environment.

Here is an example. In 2014, most money managers were under the belief that bond yields were going to rise and credit would outperform. As a result, most money managers were short-duration (relative to their benchmark) and long-credit products. This construction would have added value if the forecasts for bond yields and credit spreads had transpired in 2014. This construction did not work as bond yields fell and credit spreads widened in the second half. Most real money portfolios that were short-duration underperformed.

This is a perfect example of how the industry puts too much of its focus on forecasting bond yields and does not pay enough attention to the portfolio construction process. First, its success heavily relies on forecasting the direction of interest rates. Second, and more importantly, believing that bond yields are going to rise and credit spreads are going to narrow is betting on an improving and accelerating global economy – the short duration and long credit bets are positively correlated. If the portfolio manager’s view about the economy turns out to be correct, the portfolio benefits a lot (as bond yields usually rise and credit spreads narrow in these scenarios). On the contrary, if the portfolio manager’s view about the economy turns out to be false, as was the case in 2014, the portfolio suffers a duration and credit double-whammy. There is no diversification in this setup.

Bonds (duration risk) as a portfolio diversifier

It is our belief that in a muted inflationary environment, long bonds continue to provide portfolio diversification even as interest rates make new historical lows. In fact, diversification benefits of long nominal bonds are the best in a deflationary environment. Monitoring relative valuations between interest rates, risky assets, and the correlation environment are keys to constructing efficient portfolios that rely less on return forecasts for adding value.

This argues that long bonds do play a role in efficient portfolios even as interest rates continue to fall. That does not mean that the percentage of bonds in the portfolio should be high, low, or static. The percentage should change as the volatility and correlation environment evolves. For example, at the beginning of 2015, the environment pointed to an attractive valuation for credit products relative to rates on a risk-adjusted basis with substantial diversification benefits. Today, the relative attractiveness of credit products to risk-free bonds have normalized and the diversification benefits of having both bonds and risky assets in the portfolio have actually further increased.

These are the inputs to our portfolio construction process that we continually monitor and refine. Both long bonds and risky assets (credit in fixed-income portfolios and stocks in multi-asset mandates) continue to play a significant role in constructing efficient portfolios. Subscribing to simple arguments such as “the Fed is going to raise rates or bond yields are too low, therefore, I should reduce fixed income exposure” is likely to leave money on the table and is certainly not efficient.

And diversification is more important than ever…

Coming at it from an interest rate perspective, we see the drivers of asset prices differently. While most commentators attribute the rally in stock prices to improving growth, we see it as a forced migration to risky assets by global central banks.

Most do not remember but in 2012, for the first time, 2-Year German bund yields were negative. That was a shock to the investment management community back then but was explained away as an anomaly caused by the European debt crisis – hopefully never to happen again.

Today, the overnight rates in most countries in Europe are negative. Bond yields up to five years are close to zero or negative in most regions. What this means is that investors are paying (in nominal terms) for the privilege of lending money to governments. Looking at it another way, this is a form of capital confiscation as the investor gets less money back than was originally lent to the government. This time, however, the money management community is relatively calm about the meaning and consequences of low and negative yields.

Still, global central banks from Europe to Australia, Japan, and even Canada are easing and cutting. The question that investors should be asking themselves is this: if the global economy is fine, then why are central banks cutting interest rates?

We see this as a fragile environment that is leading to a continued melt up in risky assets. It is our view that this is not a normal business cycle and all is not fine with the global economy. This is a generational debt/deleveraging problem that we are trying to work through hopefully without a policy mistake, shock, or accident. Central banks in their effort to smooth the pain of deleveraging have boosted all asset prices and reduced the available yield to savers and investors. The available through-the-cycle (or hold to maturity) yield to savers in the world is currently negative in cash, below 2% in government bonds, 2-3% in investment grade, and about 5% in risky assets (e.g. stocks and high yield). This is the menu of options available to portfolios today and is unfortunately a list of bad choices. It is a poor list of options because the through-the-cycle (or hold to maturity) return spectrum has normally and historically been 2%, 4%, 4-5% and 8-10% for cash, government bonds, investment-grade, and stocks.

It is in this environment of fragility that tactical asset allocation and building efficient portfolios both in fixed-income and multi-asset portfolios is critical. Building an efficient fixed-income portfolio that uses duration, curve, currency, and credit levers intelligently is not simple and requires portfolio construction, deep market knowledge, and proven portfolio management and asset allocation skills. That is what we dedicate a lot of time and effort to at Signature – to evolve, improve and have a proven track record for it.


Submitted by Tim Weichel on

Thank you for a well reasoned analysis. We have been hearing forecasts of rising bond yields since the recovery after the 2000 dot com bust. 15 years of lost opportunity for those who listened to "the consensus". Even in so-called "normal" times there have been few occasions when the total return of bonds has been negative two years in a row. If we are trying to time the bond market and avoid that a negative year then we are speculators, not investors. The future is indeed unknowable.

Add new comment

We welcome your comments and questions for the Signature team and will respond as soon as possible. Please note that all comments are reviewed for their relevance to the topics discussed in the blog, and that comments may be edited.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.