CapitalTime
Articles on investing and capital management, with a quantitative focus.
Volatility and Risk-Adjusted Return
2023-02-09
One of the main goals of my PRP portfolio is to have lower volatility, and (by extension) a better risk-adjusted return. I prefer stability over highly erratic portfolio movements.
You might ask: why do I prefer less volatility? It’s more comfortable for me, and is also more compatible with occasional withdrawals out of the portfolio. These things might not be important to other investors; that’s why I suggest that other people don’t use my portfolio.
It might be helpful to study the PRP through a stress-test period, to see how it behaves. Well, lucky us! The last 3 years (from 2020-02-09 to 2023-02-09) are an interesting period:
- the COVID-19 pandemic was declared
- stocks and economic activity crashed
- corporate bonds crashed
- there was a very strong rebound
- we entered a period of very high inflation
- interest rates went up sharply
- both stocks and bonds fell hard, again
- most recently, stocks and bonds rebounded
A lot has happened in this 3 year period, so I think it’s a
useful stress-test case study. I will compare my PRP to a standard
60/40 balanced fund, using XBAL as the most comparable
portfolio. Both portfolios are bond-heavy.
Canadian PRP vs XBAL
The XBAL fund from iShares is a generic, low-fee
balanced fund which has many of the same holdings as PRP.
First, let’s examine the “drawdowns” — the % decline from a peak to a trough. Amazingly, there were two severe drawdowns in this 3 year period.
PRP had slightly better (less negative) drawdowns.
| Portfolio | 2020 Drawdown | 2022 Drawdown |
|---|---|---|
| PRP | -13.5% | -12.6% |
| XBAL | -20.9% | -16.5% |
Next, let’s look at the annualized return (compound annual growth rate) as well as a volatility measure, the standard deviation. Higher standard deviation = more volatility.
We would like to see a favourable tradeoff between risk (volatility) and return. The Sortino ratio is my preferred measure of risk-adjusted return; it places emphasis on downside volatility, just as humans do. A higher Sortino ratio = better risk-adjusted return.
PRP had less volatility and better risk-adjusted return.
| Portfolio | Annual return | Std dev | Sortino |
|---|---|---|---|
| PRP | 3.9% | 8.1% | 0.52 |
| XBAL | 3.6% | 11.5% | 0.35 |
American version of PRP
I only have high granularity (daily) data for the Canadian PRP, which is what I actually invest in. For an American version of the PRP, I’ll have to use data from the Portfolio Visualizer.
The comparison fund is Vanguard Balanced Index, which again is a low-fee balanced fund. To do a fair comparison, I changed the composition of PRP-US to use American ETFs. Here’s a link to the comparison.
I’m not posting the drawdown comparison, because the coarse granularity data at Portfolio Visualizer (monthly prices) under-states the actual drawdown. PRP-US had milder drawdowns, but it’s hard to say by how much.
PRP-US had less volatility, but almost identical risk-adjusted return.
| Portfolio | Annual return | Std dev | Sortino |
|---|---|---|---|
| PRP-US | 2.4% | 9.5% | 0.32 |
| Vanguard Balanced | 3.3% | 14.3% | 0.34 |
Note that the PRP-US has different underlying holdings than the Canadian PRP. Additionally, the Canadian comparison used 3 years of data to today, whereas the American comparison used 3 years ending 2022-12-31.
Conclusions
Over the last 3 years (starting just before the pandemic), my Canadian PRP portfolio had lower volatility and a better risk-adjusted return than a standard 60/40 “balanced” portfolio.
The American PRP also had lower volatility than a standard 60/40 portfolio, but virtually the same risk-adjusted return.
In both cases, the PRP also had milder drawdowns. I think it’s successfully doing its job as a conservative portfolio with lower volatility than 60/40.
— Jem Berkes