Trend-filtered DCA made $350,000 less than standard DCA. So why does the Calmar Ratio say it is the better strategy?

This backtest ran two approaches to investing $500 per month in the S&P 500 over a full market cycle. Same capital. Same index. Same monthly injection amount. The only difference is what each strategy does when the market turns bearish — and that single difference changes everything about whether you can actually follow the strategy through a real crash.

How Each Strategy Works

Standard DCATrend-Filtered DCA
Injection$500 on the 1st of every month$500 saved per month — not always deployed
Bear market ruleStay invested — no exceptions—
During a crashRide the drop, keep buying—
Re-entryNever needed — never exits—
FilterNoneMonthly EMA(10) crossover
Bull signal—Price above EMA(10) → invest + add monthly savings
Bear signal—Price below EMA(10) → sell 100%, hold cash
Re-entry mechanic—First bullish month → deploy ALL accumulated cash as lump sum

The critical mechanic in the Trend-Filtered approach is not just pausing — it is a full exit. When the monthly close drops below the 10-period EMA, the strategy sells the entire portfolio and converts to cash. During the bearish period, $500/month keeps accumulating in cash. When EMA(10) turns bullish again, everything — existing cash plus all saved monthly injections — goes in as one lump-sum purchase.

flowchart TD A([Start of month]) --> B{Monthly close above EMA(10)?} B -- Yes — Bullish --> C{Currently in cash?} B -- No — Bearish --> G[Sell 100% of holdings → all cash] C -- No — Already invested --> D[Add $500 to existing position] C -- Yes — Re-entry --> E[Deploy ALL accumulated cash as single lump sum] G --> F[Hold cash + save $500 this month] D & E & F --> H([End of month — repeat])

The lump-sum re-entry is what recovers part of the return gap. When a bear market ends, the first months of recovery are typically the fastest. Deploying all accumulated cash at once captures more of that early upswing than trickling it back via DCA would.

The Backtest Results

Both strategies ran on identical S&P 500 data with $500 monthly injections and $200,000 in total capital deployed.

$1,201,573
Final Value — Trend-Filtered
$1,551,685
Final Value — Standard DCA
$350,112 — Standard DCA wins on raw return
Return Gap
5.56%
CAGR — Trend-Filtered
6.38%
CAGR — Standard DCA
-18.87%
Max Drawdown — Trend-Filtered
-52.49%
Max Drawdown — Standard DCA
0.2946
Calmar Ratio — Trend-Filtered
0.1215
Calmar Ratio — Standard DCA

Standard DCA wins the return competition by $350,000. Trend-Filtered DCA wins every risk metric — and by a wide margin on Calmar Ratio.

Why Calmar Ratio Matters More Than Final Value

The Calmar Ratio puts return and risk on the same scale:

Higher is better. A higher Calmar means you are earning more annualized return for each unit of maximum pain absorbed. By this measure, Trend-Filtered DCA is 2.4× more efficient than Standard DCA.

Trend-Filtered DCA Standard DCA
CAGR 5.56% 6.38%
Max Drawdown -18.87% -52.49%
Calmar Ratio 0.2946 0.1215
Calmar Advantage 2.4× higher —

The problem with ranking strategies by final value alone is that it ignores whether a human being can actually hold through the drawdown required to earn that return. A 6.38% CAGR that demands you watch your portfolio fall 52% is not achievable by most investors — because most investors sell somewhere around -30% to -40%.

💡 TIP
The Calmar Ratio measures the quality of a return, not just its size. Think of it as: "for every 1% of max pain I accepted, how much annual return did I get?" Standard DCA gives you 0.12% of return per 1% of max drawdown. Trend-Filtered gives you 0.29% — nearly three times the return quality per unit of risk absorbed.

The -52% Problem: What a Crash Actually Feels Like

Apply the drawdown numbers to a real portfolio at the $200,000 mark.

Standard DCA — at max drawdownTrend-Filtered DCA — at max drawdown
Portfolio before crash~$200,000~$200,000
Portfolio during -52% crash~$95,000—
Paper loss in dollars~$105,000 gone~$38,000 on paper
Years of $500/month injections destroyed~17 years of savings—
Psychological stateBeyond most people's pain thresholdUncomfortable but survivable
Typical outcomeSell at loss, abandon strategyHold strategy, redeploy at next bullish signal
Portfolio during -19% drop—~$162,000
Cash accumulated on sidelines—Ready to deploy at re-entry signal

The 2008–2009 financial crisis dropped the S&P 500 roughly 57% from peak to trough. Investors who held through that entire move were eventually rewarded with substantial gains. But behavioral finance research consistently shows that most retail investors — not the weak ones, average ones — exit somewhere between -30% and -50%, precisely the zone where Standard DCA requires maximum conviction.

🚨 DANGER
A strategy that theoretically returns 6.38% CAGR only delivers that return if you hold through a 52% collapse without selling. If you exit at -40% — which most people do — your realized return is lower than what Trend-Filtered DCA produces, despite the lower theoretical CAGR. The backtest number is not your number unless you actually hold through it.

The Trend-Filtered approach changes the worst moment from "watch $105,000 evaporate" to "watch $38,000 fall on paper while cash sits ready on the side." Those are not the same emotional experience.

The $350,000 Trade-Off

Standard DCA produced $350,112 more at the end of the backtest. Trend-Filtered DCA avoided a 52% crash and came out with a 2.4× better Calmar Ratio.

Neither strategy is wrong. The right choice depends entirely on who you are.

You have a 25-year horizon. You will not check the portfolio during crashes. You have stable income independent of the portfolio, so a 52% paper loss does not change your daily life. You understand intellectually and emotionally that crashes are temporary.

Strategy: Standard DCA. Take the extra $350k. The backtest says it wins, and if you can actually hold, it does.

The requirement: You must hold through -52%. Not just in theory — in practice, at the moment it happens, when everyone around you is selling and the news says the world is ending. If you sell even once, your realized return drops below Trend-Filtered.

How the EMA(10) Filter Trades in Practice

EMA(10) on monthly bars smooths roughly 10 months of price data. It catches sustained trend reversals while filtering most short-term volatility.

Signal rules:

  • Close above EMA(10): bullish — stay invested, add monthly injection
  • Close below EMA(10): bearish — sell 100%, hold cash + continue saving $500/month
  • First close back above EMA(10): re-entry — deploy ALL accumulated cash as a single purchase

The filter will occasionally generate false signals — a brief dip below EMA(10) followed by rapid recovery. These whipsaws are the cost of the system. Over a full market cycle, they are outweighed by the protection during genuine sustained bear markets.

ℹ️ INFO
EMA(10) monthly is not a precise entry tool. There can be 1–2 months of lag between a bear market beginning and the EMA crossover trigger. Re-entry may also come 1–3 months after the actual market bottom. The strategy accepts this imprecision in exchange for reliable drawdown reduction during full bear cycles — not perfect timing, just survivable timing.

This approach has an academic foundation. Meb Faber's 2007 paper A Quantitative Approach to Tactical Asset Allocation tested SMA(10) monthly on the S&P 500 and found nearly identical results — similar drawdown reduction, slight CAGR cost, significantly better risk-adjusted returns. Faber used a simple moving average rather than exponential; SMA(10) weights all 10 months equally and tends to generate slightly fewer whipsaws than EMA(10), which front-weights recent data. Both are valid — EMA reacts marginally faster to trend changes, SMA is more stable. For a monthly DCA system, the difference is small.

⚠️ WARNING
This strategy is designed for tax-sheltered accounts — IRA, 401(k), or equivalent. In a taxable brokerage account, every EMA(10) bearish crossover triggers a sell event and a capital gains tax liability. Depending on your marginal rate and how long you held the position, tax drag could erase a significant portion of the Calmar advantage. Run the numbers with your tax rate before implementing in a taxable account. In a tax-sheltered account, this concern disappears entirely.

Your Behavioral Break-Even

⛶ Full Screen

The calculator above adds the behavioral layer that the raw backtest misses. Enter your personal pain tolerance — the drawdown percentage at which you would realistically consider selling. At any pain tolerance below 52%, Standard DCA's theoretical return advantage shrinks or disappears entirely, because you would exit before capturing the full recovery.

The One Sentence That Decides

Trend-filtered DCA trades 0.82% in annual return for the mathematical certainty that your worst moment will be -19%, not -52%.

Whether that trade is worth it is not a quant question. It is a question about what you actually know about yourself — not in a calm market, but at the moment when your portfolio reads $95,000 and the news says it is going to $50,000.

Most people overestimate how much pain they can hold through. Trend-Filtered DCA is the strategy built for that reality.