### Introduction It may seem this type of investment strategy has been around forever, with references to it abound, from academic journals to posts on social media. The rules are simple: 60% of investments go into stocks and 40% into bonds with periodic rebalancing (typically done quarterly or annually). The Wellington fund is considered to be the first in the U.S. to implement a strategy of holding roughly 2/3 of the capital in stocks and 1/3 in bonds way back in 1929, thus giving investors an opportunity to bet their money on the idea of "don't put all your eggs in one basket." Splitting capital among different opportunities was common sense and was known for millennia, but it wasn't until 1952 that Harry Markowitz showed in his dissertation paper the benefits of diversification within a mathematical framework. The conclusion was clear: when selecting assets to invest in, one should consider expected returns, risks, and assets' similarity. By combining assets that are not closely linked (but both offer positive returns over the long term), one can get the same rate of return with less risk. The good performance of the 60/40 model made it widely adopted by institutions and a golden benchmark against which many active strategies are measured. During the COVID crisis, there were debates about whether 60/40 became inadequate or just experienced a temporary setback. But a subsequent strong stock market rebound made skeptics sound like doom prophets. This strategy may be simple, time-tested, and profitable. However, it's still helpful to look inside it and "live through" the investment process. Given the democratization of financial data and technology access, let's model it.   ### Historical Simulation Each asset will be represented by a corresponding index (S&P 500 for stocks and S&P U.S. Treasury Bond Index for bonds). We choose indices over ETFs to have longer price histories. The study period is from 1990 to 2024 inclusive. Some caveats to keep in mind: 1. Unlike ETFs/mutual funds, indices don't have expense ratios (operational costs for running a company). 2. Indices have dividends built in. 3. Transactional costs when rebalancing are not taken into account. 4. No taxes on capital gains at the end of the year. The absence of real-life friction skews the results higher, but not as much as to invalidate the whole analysis. Assume it's the first business day of 1990, and an investor has $100,000 of capital. He buys $60,000 worth of stocks and $40,000 worth of bonds. On every first business day of the new quarter, he trades if necessary to bring allocation in line with the 60/40 percent target mix. First, some basic stats: | Period | 1990-2024 | | :--------------- | --------: | | Starting Capital | 100,000 | | Ending Capital | 1,793,811 | | CAGR | 5.9 | | Max Drawdown | -32.3 | | Profitable Years | 29 | | Losing Years | 6 | | Asset | Percent of Total Profit | | :----- | ----------------------: | | bonds | 11.7 | | stocks | 88.3 | Even though we targeted a 60/40 allocation mix, 88% of the total profit came from stocks. ![[attachments/60_40_nav.svg]] ![[attachments/60-40-returns-by-year.svg]] Looking at the equity curve over long time horizons may create a false impression of an easy ride. However, those who traded this strategy from 2000 to 2010 would tell a different story. It was a decade of a barely positive annualized rate of return and violent drawdowns. Two of the deepest drawdowns happened during this period. | Start | Lowest Point | Recovery | Days | Max Drawdown | | :--------- | :----------- | :--------- | ---: | -----------: | | 2007-10-10 | 2009-03-09 | 2010-04-22 | 926 | -32.3 | | 2000-09-05 | 2002-07-23 | 2004-11-02 | 1520 | -25.9 | | 2020-02-20 | 2020-03-23 | 2020-06-04 | 106 | -20.2 | | 2021-12-28 | 2022-10-14 | 2024-01-19 | 753 | -19.3 | | 1990-07-17 | 1990-10-11 | 1991-02-01 | 200 | -11.8 | Let's take a closer look at stressful periods. We see that the stock component of the portfolio was responsible for the decline of the portfolio value for the most part. In 3 out of 4 displayed drawdowns, bonds acted like a counterbalance, providing some protection to the portfolio. However, this relationship broke down during the COVID crisis when both assets tanked. ![[attachments/60-40-drawdowns.svg]] The [[Glossary#^correlation|correlation]] between assets' returns for this period was -0.01, virtually 0. A more human-friendly way to look at it is to make a scatter plot with an x-axis displaying monthly return for stocks and a y-axis for bonds: ![[attachments/60-40-stocks-bonds-returns-scatter.svg]] We see that points form a cloud without any particular pattern, meaning assets' month-to-month price fluctuations are more or less independent of each other. Sure, there are cases when both stocks and bonds went up in price, but likewise, there are ones when both assets tanked simultaneously. In general, though, their prices zig-zag on their own. ### Summary 1. The 60/40 model has a long track record and reasonably strong results due to the exceptional performance of the U.S. stocks. 2. The portfolio dynamics is mainly determined by the stock component, both in good and bad times. 3. The model is easy to implement with low-cost ETFs and doesn't require much maintenance. 4. There may be prolonged periods of poor performance, which can test investors' ability to stick to this method. 5. Main question: is there room for meaningful improvement that justifies the effort?