Stocks have higher historical and expected returns than bonds, but they are highly volatile.
The benefit from adding more stocks declines and approaches zero at (and can even turn negative) at high levels of equity concentration due to volatility and compounding.
Having an all stock portfolio increases the potential range of portfolio outcomes, and it is not guaranteed to be the best even over long time periods.
Stocks vs. Portfolios Over the Long Term
Stocks have a higher historical and expected returns than bonds, even risky long term bonds. But, because of compounding, adding to diversifying assets can add value even for a return-maximizing investor.
Everybody knows that stocks outperform bonds over the long term. In the chart below, we’ve constructed portfolios that are combinations of long-term bonds, which are very risky, and stocks, starting with all bonds at the left to all stocks on the right, and mixes of the two in between. Looking at the returns of the assets, and average monthly returns of the portfolios, the lesson seems clear:, the more stocks you have, the higher your return. In the charts below, we show the returns over the last 50 years of these portfolios. In the table, excess returns means total returns less the returns from holding in cash, in order to measure the returns from putting money at risk in the asset over the time period).
However, as investors, we don’t care about average monthly returns. We care about performance over longer periods of time. So let’s look at the compounded returns of our portfolios, not just the average monthly returns. The picture changes. The benefit of owning stocks declines significantly after 60% stocks, and rounds to zero after 80% stocks. Because diversification reduces volatility and having lower volatility helps compound returns, the benefit from owning stocks declines as the stock concentration of the portfolio increases.
Another way to look at it is in terms of return to risk ratio. The return to risk ratio peaks around 50-70% in stocks. After that point, the portfolio gains less and less from adding stocks and taking on more risk.
The main point is that the benefit of taking on more stocks declines as stock concentration increases, and flattens after the portfolio reaches 70-80% stocks. Assets that provide diversifying returns help portfolios significantly. This is why we believe it is better to have some diversifying assets like government bonds and gold compared with owning more stocks for growth investors. Even if you reduce your return expectations for bonds below the returns from the past fifty years, the basic dynamics hold: each marginal dollar of stocks has less benefit because of compounding and diversification.
These general patterns are consistent with the math of how portfolios and compounding work, not simply one case study. They also hold up under stress tests with lower returns from bonds and different correlations between bonds and stocks. There are good arguments that bonds will underperform in the next 20 years (though those kinds of arguments, and we, are often wrong). There are also good theoretical reasons that stocks might offer better returns per unit of risk than bonds. Bonds often offer high returns in hard economic times, and stocks perform poorly in those same hard times, right when people might want to draw on savings. Because investment returns are compensation for taking on risk, it makes sense that the asset with insurance-like qualities would have lower returns.
Over Shorter Time Horizons the Distribution of Returns Matters
However, in the real world, most investors don’t have 50 years. They have maybe 40 year careers, and they add a lot of their lifetime savings as they get older and salaries get higher, meaning that a lot of the money they save does not have 40 years to compound. Also, investors don’t experience average returns, they experience returns that are specific to a given time period.
So, in addition to thinking about the highest expected returns, investors are well served by considering the potential range of outcomes over time periods that are relevant to them.
Going back to the same simulation as before, we calculated the range of returns that investors would have experienced over 5 year periods. We use 5 year periods so that our samples don’t overlap too much (there are only 2.5 20 year periods in 50 years, and long term bond data is not good before the 1970s). The dynamics over 5 years are the same as for the full time period. Performance improves as we add stocks to the portfolio, but the impact declines as we near 100% stocks. Note that we’ve used returns in excess of the cash rate to enable a fair comparison across time periods. At the lower range of outcomes, stocks can perform poorly relative to more diversified portfolios, as can long term bonds.
So, should you allocate to all stocks? It depends.
At Wealthsimple, we advise investors with long time horizons to hold a growth portfolio with 80% stocks and risky bonds (+ gold) because the benefit of allocating to more stocks decreases as we go from 80 to 100% stocks, and we value improving the worst potential outcomes. We don’t think in terms of money, we think in terms of our investors’ life outcomes, and we value making sure they receive adequate returns in bad scenarios over achieving the absolute best returns when equity markets are booming. Our investors will do very well holding 80% equities in that case anyway. For our more conservative portfolios, we add shorter term bonds instead of long term bonds because a heavy allocation to risky bonds can also lead to bad outcomes. Those portfolios are not shown here.
However, investors differ, and uses for money differ, so it’s a matter of determining the right tradeoff for your risk appetite and time horizon. If you want a chance at the highest possible returns, an all stock portfolio is probably right. If you care about the distribution, and making sure you are compensated well for taking on risk even over a long time horizon, it may make sense to allocate some of your portfolio to risky government bonds.
Disclosures: The statements provided here are for information and educational purposes only and does not constitute advice or a recommendation. The indicated performance are historical for the period indicated. The rate of return does not take into account any fees or tax payable. Past performance may not be repeated. Portfolios consist of the MSCI World Total Return Index and the U.S. 30 Year Treasury Total Return Index, rebalanced monthly, at different equity and bond rates, from 1970-2021. Cash rates used in excess returns are the Global Financial Data T-Bill Total Return series. Excess returns are calculated based on compounded total returns less compounded cash returns. Return-to-risk ratio defined as excess returns divided by annualized monthly standard deviation of returns. Data from Global Financial Data. Analysis by Wealthsimple.
