Portfolio Allocation Optimizer

The replay uses real (partly synthetic-backfilled) monthly adjusted prices — dividends and splits included; trading costs, taxes, spreads and FX are not. Prices come from Yahoo Finance via automated processing and may contain errors or gaps; synthetic-backfill segments are estimates that can deviate from actual performance, so treat the results as indicative only. Past performance does not predict future returns.

Portfolio allocation — FAQ

What does this calculator actually do?

It compares two assets and simulates every capital split between them — from 100% in one to 100% in the other — across thousands of randomized market paths. For each split it reports the range of outcomes after your chosen number of years, then highlights the allocation that's best for your selected objective.

What's the difference between the three objectives?

‘Risk-adjusted’ maximizes return per unit of risk (CAGR ÷ volatility) and usually lands on a blend. ‘Calmar’ maximizes return per unit of max drawdown (CAGR ÷ drawdown) and also favors a blend. ‘Min drawdown’ minimizes the typical worst peak-to-trough fall (max drawdown) — the calmest split through a crash.

How do I know the correlation between my two assets?

You don't need a number — pick the plain-language option that fits. Two funds in the same market move almost in lockstep; stocks in different regions move mostly together; stocks versus gold are loosely linked; stocks versus bonds sometimes move oppositely. When you pick two built-in ETF presets, a sensible value is filled in automatically.

Why does a mix sometimes beat putting everything in the higher-return asset?

Because the two assets don't move in perfect lockstep, a blend has lower volatility than the average of its parts. Lower volatility means less drag from big swings, so the risk-adjusted and downside outcomes can beat a single concentrated bet — even if the median slightly favors the higher-return asset.

What does ‘rebalance yearly’ change?

With rebalancing on, each year you trim the asset that grew and top up the one that lagged, restoring your target split. This keeps risk controlled and is the standard efficient-frontier assumption. With it off, the higher-return asset gradually dominates, and over long horizons the ‘optimal’ split drifts toward simply holding the winner.

Is the expected return a CAGR or an average?

It's the CAGR (compound annual growth rate) — the median, geometric outcome. The simulation treats your input as the median path and lets volatility spread the distribution around it, consistent with how ETF returns are usually quoted.

What do P10, P50, and P90 mean?

They're percentiles of the ending wealth across all simulated paths. P50 is the median (typical) outcome, P10 is an unlucky result (only 10% of paths end lower), and P90 is a lucky one (only 10% end higher). The gap between P10 and P90 shows how much risk a given allocation carries.

Can I use this for assets other than ETFs?

Yes. The presets are a convenience that fill in a return and volatility, but you can choose ‘Custom’ for either asset and enter any CAGR and volatility — for stocks, bonds, crypto, real estate proxies, or anything you can characterize with those two numbers plus a correlation.

• Stocks vs bonds (60/40)

  Why does the classic 60/40 work? Monte-Carlo the median return and downside of different stock/bond splits to find the allocation that fits your risk.

• QQQ vs SPY

  Nasdaq 100 (QQQ) returns more but swings harder; S&P 500 (SPY) is steadier. Instead of choosing, compute the best blend and its downside risk.

• Aggressive growth mix

  Young, long horizon, can stomach big swings? Model a Nasdaq-100-plus-semiconductors growth mix over 25 years to see the upside — and the cost — of going aggressive.

• US vs international

  All-in on US stocks, or spread some abroad? Monte-Carlo the downside of concentration vs geographic diversification to find your global allocation.