Income-Scaled Monthly Investing: A Historical Counterfactual Study

Research question

If a saver’s monthly contribution scales with a growing income — rather than staying fixed in dollars — what wealth paths, drawdown experiences, and money-weighted returns would 26 assets have delivered over 2000–2026 under one identical cash-flow rule?

Method

The cash-flow rule is fixed by configuration, not fitted: initial income $5,000/month, contribution rate 30%, income growing 5% at each calendar-year start, invested on the first trading day of each month from 2000-01 through 2026-05. All results come from audited output tables with a single terminal valuation date (2026-05-29).

Two comparison bases are kept deliberately separate:

• Primary B (own inception). Each asset starts at its own first investable month — descriptive, but mixes unequal histories.
• Unified windows. Common start/end months for all assets; W2018 is the main fair comparison because every configured asset is represented in it.

Money-weighted return (XIRR) is the headline metric because contributions are the object of study; best/worst-case drawdowns of the account path accompany every return figure.

Findings

• The window picks the winner. TSLA tops the own-inception view (XIRR 43.36%) while NVDA tops the fair window W2018 (70.98%) — the same window trap familiar from price-history rankings, now operating through contribution timing.
• Late dollars dominate. Because contributions grow with income, late-cycle market behavior dominates terminal wealth — an income-scaled saver’s experience is much more exposed to the last few years than a fixed-dollar DCA intuition suggests.
• The benchmark is sober. SPY compounds 317 contributions into $4.63M at a 12.25% XIRR (own-inception), with a -52.9% worst account drawdown along the way; in W2018 it earns 17.39%. Only a handful of ex-post winners beat it decisively.
• Return buckets, with pain attached. High-return/moderate-pain assets (GOOG, XLK, WMT, GLD, SPY, MSFT) cluster around a 22% median XIRR; steady defensives (KO, XLU, XLP, XLV, PG, SHY, IEF) around 7.25%; long-duration Treasuries (TLT) lose -3.68% in W2018 in the 2018–2026 rate regime. Extreme winners (NVDA, TSLA) pair extreme XIRR with severe path drawdowns.

What this is, and is not

It is a historical counterfactual of one disciplined cash-flow rule, useful for calibrating what “investing a growing income” actually felt like — including the drawdowns. It is not a forward-looking asset-selection rule: the universe contains known ex-post winners, so the results carry hindsight selection and survivorship bias by construction, and the paper says so explicitly.