USD basis · includes currency movements Past data: 119 months, 2016.11–2026.09 · not a forecast

Best return for its risk

The mix with the most past return per unit of volatility, measured by the Sharpe ratio.

These weights

Total 100%

SPY 100.0%
TLT 0.0%

All 100% went to SPY

The mix leans on what rose most in the past. A slightly different period can change the weights a lot.

Measure

Same assets and period; the weights change with what each measure tries to get best.

Best return for its risk 15.7% 15.4% Smallest swings 5.9% 11.0% Smallest bad-month loss 6.6% 11.1% Risk shared evenly 6.6% 11.1% Best return for its downside 15.7% 15.4%

Compared with equal weight

Compared with equal weight
Metric	Equal weight	These weights	Difference
Past average annual return	7.1%	15.7%	+8.6pp
Volatility (annual)	11.1%	15.4%	+4.2pp
Max drawdown · month-end	-26.8%	-23.9%	+2.9pp
Avg. loss in worst 5% of months	-6.5%	-9.2%	-2.7pp
Sharpe ratio Target	0.42	0.86	+0.45
Sortino ratio	0.62	1.36	+0.75

Best return for its risk Equal weight

Show the 24 mixes on the frontier

Mixes on the efficient frontier and their weights
Mix	Past average annual return	Volatility (annual)	Weights
Mix 1	5.9%	11.0%	TLT 57% · SPY 43%
Mix 2	6.3%	11.1%	TLT 55% · SPY 45%
Mix 3	6.7%	11.1%	TLT 52% · SPY 48%
Mix 4	7.2%	11.1%	SPY 50% · TLT 50%
Mix 5	7.6%	11.2%	SPY 53% · TLT 47%
Mix 6	8.0%	11.3%	SPY 55% · TLT 45%
Mix 7	8.4%	11.4%	SPY 58% · TLT 42%
Mix 8	8.9%	11.5%	SPY 60% · TLT 40%
Mix 9	9.3%	11.7%	SPY 63% · TLT 37%
Mix 10	9.7%	11.8%	SPY 65% · TLT 35%
Mix 11	10.2%	12.0%	SPY 68% · TLT 32%
Mix 12	10.6%	12.2%	SPY 70% · TLT 30%
Mix 13	11.0%	12.4%	SPY 73% · TLT 27%
Mix 14	11.4%	12.6%	SPY 75% · TLT 25%
Mix 15	11.9%	12.8%	SPY 78% · TLT 22%
Mix 16	12.3%	13.1%	SPY 80% · TLT 20%
Mix 17	12.7%	13.3%	SPY 83% · TLT 17%
Mix 18	13.2%	13.6%	SPY 85% · TLT 15%
Mix 19	13.6%	13.8%	SPY 88% · TLT 12%
Mix 20	14.0%	14.1%	SPY 90% · TLT 10%
Mix 21	14.5%	14.4%	SPY 93% · TLT 7%
Mix 22	14.9%	14.7%	SPY 95% · TLT 5%
Mix 23	15.3%	15.0%	SPY 98% · TLT 2%
Mix 24	15.7%	15.4%	SPY 100%

Method and limits

The baseline and all five measures use the data and assumptions below.

Actual data period: 2016.11–2026.09
Monthly observations: 119months
Average annual return: Average monthly return × 12, so it can read higher than the backtest's annualized return
Covariance: Annualized monthly sample covariance
Weight maintenance: Monthly rebalance to target
Return scope: Dividends reinvested · 0bp transaction costs · Pretax, nominal returns
Risk-free rate: 2.46% · TB3MS
Weight constraints: 0%–100%
Feasible return range: -1.5%–15.7%

These estimates use historical data and do not guarantee future performance or prevent losses.

How to read portfolio optimization results

What does the optimizer calculate?

The optimizer uses historical monthly returns for your selected assets to estimate return and risk at different weights. It compares how allocations change when the objective changes while the assets and calculation conditions stay the same.

How the five measures differ

Best return for its risk · Max Sharpe: The mix with the most past return per unit of volatility, measured by the Sharpe ratio.
Smallest swings · Min Volatility: The mix that swung the least in the past, measured by volatility.
Smallest bad-month loss · Min CVaR: The mix with the smallest average loss in its worst 5% of months, measured by CVaR.
Risk shared evenly · Risk Parity: The mix in which each asset adds a similar share of the total swings.
Best return for its downside · Max Sortino: The mix with the most past return per unit of downside volatility, measured by the Sortino ratio.

The default baseline allocates equally across all assets. When you load a portfolio, its weights become the baseline for comparing the five optimized allocations.

Actual calculation period and observations

This result uses 119 monthly observations from 2016.11 to 2026.09. The actual calculation period can differ from the requested number of years, so check the dates shown with the result.

Data as of 2026.10.29

Return estimates and operating assumptions

Expected return and risk use annualized historical monthly means and sample covariance in USD. The calculation assumes monthly rebalancing to target weights, dividend reinvestment, and 0bp transaction costs. Taxes and inflation are excluded.

Each asset weight is between 0% and 100%, and weights sum to 100%. Changing the weight limits changes the feasible allocations and optimization results.

Limits of these results

The displayed returns and weights are estimates from historical data, not guaranteed future optimal weights or returns. Results can change with the period and asset selection. Use a backtest to examine loss periods and operating conditions for the allocation you choose.