What earns a place in a portfolio
Most things people bolt onto a portfolio look good on their own and do nothing once they are inside one. A few look unimpressive alone and help the whole. The only test that matters is the whole portfolio.
The idea behind this site is simple. No single strategy beats the market by much, and none does it reliably. But if you can find several ideas that each have a modest edge, and they do not all fail at the same time, stacking them gives you better odds than any one of them. The catch is that most candidates fail the second half of that sentence. They either add nothing once fees and taxes are counted, or they fall exactly when your stocks fall, or they only help if you buy them a particular way.
So every idea below was tested the same way: put it inside a whole portfolio, charge its costs, and compare the result with a cheap index fund holding the same mix. The verdict words are the four I use everywhere on this site. Settled means the direction is not in doubt. Probably means the evidence points one way but a long stretch of bad luck could hide it for decades. Too close to call means the test could not separate the idea from zero. No means it lost, on measured cost against measured benefit.
The verdicts
| Idea | Verdict | In one line |
|---|---|---|
| Cheaper funds | Settled | Index funds cost 0.09% a year against 0.57% for the average active fund, and 90% of active US stock funds lost to their benchmark over ten years. |
| The right account | Settled | Which fund sits in your Roth, your IRA and your taxable account is worth a few hundredths of a point a year, every year, for the cost of a form. |
| An ETF instead of a mutual fund | Settled | In 2025, 7% of ETFs paid out a taxable capital gain against 52% of mutual funds. |
| Never trading | Settled | Every sale in a taxable account brings a tax bill forward. Not selling is the one free edge. |
| Leaning toward cheaper companies | Probably | Beat a plain index mix by about 0.8 points a year between 1990 and 2026, and by 0.9 points on data from the 1980s the funds were not built on. Needs thirty years of patience. |
| Holding foreign stocks | Probably | Worth holding because foreign markets are priced at a big discount to the US, and because nobody knows which region wins the next twenty years. Not because it protects you in a crash. |
| Trend following, on top of stocks | Too close to call | The strongest candidate on this site. Added 2 points a year over 96 years, most of it before 2009, and the fund that does it is under three years old. |
| The 200-day rule | No | Cuts the worst fall from 84% to 43% and pays for it with 0.7 points a year less than holding, 58 losing exits out of 73, and a tax bill. |
| 2x and 3x funds | No | A 3x fund held since 1926 spent 28 years under water and fell 99.9% at its worst. Under the 200-day rule it still fell 85%. |
| Gold | Too close to call | Earned 1.75% a year over cash since 1975 and rose in 59% of the worst stock months since 2000. Small, optional, and instead of cash rather than instead of stocks. |
| Bitcoin | No | The only holding tested that made the portfolio's worst fall deeper at every weight. A bet, not protection. Zero to 2% if you want the bet. |
| Crash insurance and buffer funds | No | The best-known tail fund lost 8% a year while stocks made 14%, and fell 13% in the 2022 bear market. 86 of 102 buffer funds returned less than a plain stock and cash mix. |
| Bonds and TIPS | Too close to call | Owning fewer stocks is a choice about how much you can lose, not a way to earn more. Ten-year TIPS, Treasury bonds whose payments rise with inflation, now pay 2.4% above inflation, the most since 2008. |
| Cash | Settled | A 10% move from stocks into Treasury bills added 0.92% back in the average worst stock month since 1926, as much as gold and within two hundredths of long bonds, for no fee. |
| Corporate and high-yield bonds | No | Investment-grade corporate bonds track Treasuries at 0.83 out of 1, and high-yield rose in 3% of the worst stock months. A little extra return inside a bond allocation, never protection. |
| Foreign and emerging-market bonds | Too close to call | Not tested here. A hedged foreign bond fund is close to BND with a currency hedge attached; emerging-market bonds carry the currencies that fall hardest in a crisis. |
| Catastrophe bonds | Too close to call | The one asset whose losses come from hurricanes. The funds returned about one point a year over cash after fees, and the price is the thinnest on record. Wait. |
| Commodities | No | Gained 144% across 1973 and 1974 and fell with stocks in most crashes, rising in 28% of the worst stock months since 2000. An inflation hedge, not a crash hedge. |
| Carry | Too close to call | A second strategy run with futures contracts, which moves independently of trend. Positive on paper, negative since 2013, and the fund is two years old. |
| Direct indexing | No | Cost money on every one of 400 simulated thirty-year paths. Only 0.2% of the losses it harvests ever save tax for someone who never sells. |
| Rebalancing | Settled | Keeps your mix from drifting. Does not earn a return, and every rule tested on a three-region mix lost a little to never trading. |
| Dividend and property funds | No | A dividend fund is the stock market with a screen and a tax bill. Property funds fell 41% in 2020 and rose in 3% of the worst stock months. Quality funds lost more than the market in 2022. |
| Paying off debt, and the HSA | Settled | A loan above the after-tax Treasury rate is the best risk-free return a household can get, and the health savings account is the only account untaxed at every point. |
Alone and together are different questions
Some ideas fail alone and help in a portfolio. Trend following is the clearest case. On its own it earned about 5% a year since 2000 by the SG Trend index's count, with long dead stretches. Bought by selling stocks, it dilutes the portfolio. Added on top of stocks through a fund that borrows inside itself, it rose in 64% of the worst stock months since 2000 and turned the lost decade of 1999 to 2009 from a loss into roughly zero. Cash is the other case. Cash earns the least of anything, and moving 10% of a stock portfolio into Treasury bills would have added 0.9% in the average terrible month since 1926, as much as gold or long bonds and for no fee.
Others look strong alone and fail inside a portfolio. Bitcoin has the best return for its risk of anything I measured, and it still deepened the portfolio's worst fall at every weight because it falls with stocks and falls further. A standalone managed-futures fund, a trend-following fund, bought by selling stocks gives up the stock return on the money you sold, which is most of what the strategy was supposed to add. Gold has the same problem: bought by selling stocks it costs the portfolio 0.4 points a year against holding the stocks, added on top it gains 0.2, and neither number can be told from zero.
Ask two questions of any idea. Does it fall when stocks fall? And how am I paying for it: by selling stocks, or by adding it on top?
Numbers as of 2026-09-02. Corrections lists anything that changed.