Blended Strategies

Why not just one strategy?

• Concentration. When the market is overheated and good companies are scarce, an individual strategy may select only a few stocks. The year’s result then depends on the success or failure of one or two companies.
• Different behavior. AVI strategies select stocks using different criteria, so they do not move in sync. The correlation between their monthly returns ranges from 0.37 to 0.84, averaging around 0.67.

Even a simple equal-⁠weight blend of all eight strategies (rebalanced once a year) delivered an average annual return of 18.7% over 2000-⁠2026 - higher than the average for its constituent strategies (17.9%). Its volatility was 17.9%, compared with an average of 21.6% for the individual strategies. We publish the strategies’ monthly results in our track record.

But there is a caveat worth acknowledging. The blend’s maximum drawdown was 39%, only slightly below the average for the individual strategies (42%). The reason is clear: all AVI strategies buy shares in US companies, and in a serious crisis they fall together. In 2008, every strategy except DIVIDEND lost between 14% and 45%. Blending AVI strategies smooths out normal fluctuations but offers little protection against major crashes.

That means we need to look beyond the stock market for a counterbalance.

What to balance them with: gold and Treasury bills

We need assets that move independently of stocks. To make a fair comparison, we used the same data for all calculations: monthly returns from January 2000 to December 2025, a total of 312 months.

Gold’s correlation with Treasury bills is 0.11. Both assets are virtually uncorrelated with the market, with AVI strategies and with each other. But zero correlation alone does not provide protection. What matters is how an asset behaves specifically when stocks fall.

Treasury bills did not produce a loss in any of these episodes: investors who hold a bill to maturity know in advance how much they will receive. They protect a portfolio not by rising in price, but by preserving value when everything else falls. Gold rose in 2000-⁠2002 and 2008, and was almost unchanged in 2022. But at the height of a panic, investors also sell gold to raise cash: in October 2008, it fell by 16%, almost as much as stocks.

Gold’s drawbacks

• It is not a stable asset. Gold’s volatility is even slightly higher than that of the stock index (16.1% versus 15.2%), and its maximum drawdown was 43%: after its 2011 peak, it declined for several years.
• It generates no income. Gold has no earnings or dividends; its entire return comes from changes in price.
• Our period was favorable for gold. Gold was cheap in 2000, and its annual return of 10.5% over 26 years should not be extrapolated into the future.

Gold is therefore more of an independent source of return than an anchor. The real anchor in this pair is Treasury bills. Their drawback is different: returns after inflation are close to zero, and negative in some years. Treasury bills dampen fluctuations but do little to grow capital.

The barbell

The idea resembles a barbell strategy: at one end are assets that generate returns but are volatile; at the other are assets that support the portfolio when markets turn turbulent. There is nothing in between - no long-⁠term bonds or index fund. At the return-generating end, we have AVI strategies focused primarily on growth: AVICORE, POWER, EMERALD, INDIGO, YELLOW. At the defensive end are gold and Treasury bills.

The first and simplest test: 70% split equally among five growth strategies, 15% in gold and 15% in Treasury bills, rebalanced once a year.

Reducing the drawdown from 40% to 28% cost approximately 3.9 percentage points of annual return. Even so, the portfolio’s return remained far higher than the market’s, while its volatility was actually lower. These weights have not yet been fine-⁠tuned; this is just a rough equal-⁠weight blend. In the next article, we will show how to fine-⁠tune the allocations and how much the return-⁠risk trade-⁠off can be improved.

You can try building your own blend on the “Create a Strategy” page. Set the allocations to strategies, gold and Treasury bills, the testing period and the starting capital - the calculator will show the return, volatility and drawdown. If you cannot find a combination with your desired return and a drawdown you can tolerate, that is a valuable result too: it means you should reconsider your portfolio requirements.

Caveats

• All calculations use the same historical period, 2000-⁠2026. This is a backtest: it does not account for taxes, commissions or slippage, and it does not guarantee the same results in the future.
• The S&P 500 Total Return series is based on closing prices on the last trading day of each month.
• The GOLD and T-⁠BILLS series use the GLD fund (from December 2004) and the BIL fund (from June 2007), representing the returns an investor would actually have earned in these funds after their expenses. For earlier years, proxies are used: the Perth Mint gold price less GLD expenses (0.40% per year) and the ICE BofA 3-⁠month Treasury bill index less BIL expenses (0.14% per year). In years with near-⁠zero interest rates, BIL sometimes lost up to 0.23% in a month because of fund expenses.

The past does not guarantee the future, but it is not irrelevant to it either. In the next articles, we will show how to estimate the range of possible outcomes for a blended portfolio, rather than just its average result.

Past performance does not guarantee future results. This material is for information only and is not investment advice.