This is the last notebook in my TiPortfolio series, and honestly it is the most practical one. All the previous strategies — VIX regime switching, volatility targeting, beta neutrality — are interesting but require active monitoring and some comfort with complexity. This one is different.
Auto Investment Plan (AIP), or dollar-cost averaging (DCA), is simple: put a fixed amount of money in every month, regardless of what the market is doing.
The Strategy
The portfolio is equal-weight across QQQ, BIL, and GLD, rebalanced monthly. The difference from earlier notebooks is run_aip() instead of run() — it injects $1,000 into the portfolio at the end of each month before the rebalancing fires.
portfolio = ti.Portfolio(
"aip_equal_weight",
[
ti.Signal.Monthly(),
ti.Select.All(),
ti.Weigh.Equally(),
ti.Action.Rebalance(),
],
TICKERS,
)
result_aip = ti.run_aip(
ti.Backtest(portfolio, data),
monthly_aip_amount=1000
)That is the whole strategy. One line to run the backtest.
Results (2019–2024)
sharpe 2.433
calmar 4.494
sortino 4.862
max_drawdown -0.105
cagr 0.470
total_return 13.736
final_value 147355.743
total_fee 4.451
total_contributions 84000
contribution_count 84Starting with $10,000 and contributing $1,000 per month for 84 months ($84,000 total contributed), the portfolio grew to $147,355. That is a 47% CAGR on the portfolio value — though of course, that number is inflated by the ongoing contributions. The real metric is that you put in $94,000 total ($10k start + $84k contributions) and ended up with $147k — roughly 57% gain on total capital deployed.
The Sharpe ratio of 2.433 and Sortino of 4.862 look astronomical, but this is partly a mathematical artifact of DCA — the regular contributions smooth the equity curve significantly, making the return-to-risk ratios look much better than a lump-sum strategy.
Lump-Sum vs DCA Comparison
Metric Lump-Sum AIP
-------------------------------------------------
final_value 19951.667 147355.743
cagr 0.104 0.470
sharpe 0.655 2.433
sortino 0.875 4.862
max_drawdown -0.162 -0.105
total_return 0.995 13.736
AIP total_contributions: $84,000
AIP contribution_count: 84The lump-sum portfolio (same $10k starting capital, no further contributions) ended at $19,951. The AIP portfolio ended at $147,355. But that is not a fair apples-to-apples comparison — the AIP put in $84,000 more.
The more honest comparison is on risk metrics: AIP had a lower max drawdown (-10.5% vs -16.2%) because monthly contributions during downturns effectively average down your entry price. In the 2020 crash and 2022 selloff, the $1,000 monthly injections were buying at cheaper prices, which reduced the portfolio’s drawdown severity.
The best thing about AIP is not the returns — it is the behaviour it enforces. When markets crash, you keep buying. Most people panic and stop. The strategy removes that decision from the equation.
The DCA vs Lump-Sum Debate
Lump-sum investing statistically outperforms DCA about 2/3 of the time in rising markets, because your money is in the market longer. But DCA wins in down or choppy markets, and more importantly, it is what most working people actually have available — you invest from your salary, not a lump sum.
From what I see in the data, the real advantage of DCA for regular savers is not mathematical — it is psychological. You never have to worry about investing at the top. You just invest every month.
My Take After Testing All Six Strategies
Looking back at the whole series:
*AIP Sharpe is not comparable to the others due to contribution effects.
VIX regime switching was the most interesting for a lump-sum portfolio. But for most people, the AIP approach is the most realistic — it fits how salaries work, it removes market timing decisions, and the compounding over time is substantial.
I plan to keep adding strategies to TiPortfolio. If you have suggestions for what to backtest next, let me know.