Personal Finance (24): Is Your Trading System Really Good Enough To Trade Full-Time? (1/2)

What Returns Can a Trading System Realistically Earn? You may think your trading system is performing well. But compared with what?  

This article is a small research project using multiple AI tools to dig through publicly available and verifiable sources, looking for real-world benchmarks for trading performance.

The goal is not to sell a dream, but to establish realistic expectations—especially for new traders, who are often tempted by promises of extraordinary returns.

Before you chase the rainbow, it may be worth knowing how far away it really is.

A Sourced Benchmark Comparison of Systematic Strategies

For new market participants — what the data actually says, versus what the courses advertise

“Know the truth, and the truth will set you free.”

Because of false advertising, too many traders spend years chasing rainbows. The cure is a real number with a real source next to it.

Every bull market brings a wave of new traders and a matching wave of advertising promising returns that multiply the market — “10x the S&P,” “100% a year.” This article puts real, cited numbers next to those claims. The core message is simple: even the largest professional funds struggle to beat a plain S&P 500 index fund over the long run, and the honest return profiles of systematic strategies are far more modest — and far more useful to know — than the marketing suggests.

A note on rigour, because it matters here: every headline number is tied to a named, public source (see Section 2a), with a link you can verify yourself. Each strategy is also tagged by category — whether it is a real investable index, a professional benchmark, an academic factor, or a strategy family with no public benchmark at all. For that last kind, this article deliberately prints no return figure, because any single number would be invented — and inventing impressive numbers is exactly what we are warning you against. Do not take my word for any of it; check the sources.

1.  The Uncomfortable Baseline: Most Professionals Don’t Beat the Index

Before comparing strategies, anchor on this. S&P Dow Jones Indices publishes the SPIVA Scorecard twice a year, measuring active US funds against their benchmarks. The Year-End 2025 edition is the newest data, and the verdict has been consistent for over two decades. But note the key move: the story is not the single bad year — it is the long horizon.

It is not hard to beat the market for one year; roughly a third of funds manage it in any given year. The hard part — the part that matters — is doing it consistently over 10, 15 and 20 years. Over long windows the failure rate climbs toward 90%+:

Source: SPIVA U.S. Year-End 2025 Scorecard, S&P Dow Jones Indices (free). 79% underperformed in 2025 — worse than 65% in 2024, and the fourth-worst year in the scorecard’s 25-year history. Over 15 years, no US equity category had a majority of managers beat its benchmark.

•    The persistence problem is even more damning. S&P’s companion Persistence Scorecard tracks whether the funds that do win keep winning. Of top-half large-cap funds in 2021, only a handful stayed in the top half over the next four years — fewer than pure chance would predict. Past outperformance did not survive; when it happened, it was mostly luck, not repeatable skill.

Source: SPIVA U.S. Persistence Scorecard, S&P DJI.  Corroborated by Morningstar’s Active/Passive Barometer and Fama & French (2010), “Luck versus Skill in the Cross-Section of Mutual Fund Returns.”

Why this is the right starting point It isn’t difficult to find a professional fund that beat the market for one year. The difficult part is doing it consistently for 10, 15 or 20 years — and the data says almost none do. If full-time professionals with research teams, data budgets and direct market access mostly fail at this, any advertisement promising an untrained beginner several times the market return is, on its face, implausible. The realistic goal is not to “beat the market by multiples” — it is to earn a comparable long-run return with smaller, shorter drawdowns, so you actually survive to compound. That is achievable. The fantasy is not.

2.  Master Comparison Table

The most important thing to understand before reading this table: these rows are not all the same kind of thing. One is an investable index, one is a professional managed-futures benchmark, two are academic research factors, and one has no public benchmark at all. Mixing them as if they were directly comparable is the single biggest mistake in articles like this — so the Category column tells you exactly what each number is and how much weight it deserves.

Read the table by shape, not by any single cell. Sources for every figure are in Section 2a.

Strategy	Category	CAGR (net, full-cycle)	Sharpe	Max drawdown	Skew	Typical hold
Trend following	A — pro CTA index	5% (2000–’25) †	0.5–0.6	20% (index)	Positive	2–4 months
Cross-sectional momentum	B — academic factor	Market + a few % (gross) ‡	0.4–0.6	30–73% (factor)	Mild positive	1–6 months
Value / fundamental	B — academic factor	Market, long-run (gross)	0.3–0.6	20–35%	Neutral	1–3 years
Mean reversion (short-term)	C — no public index	No defensible public figure — must be individually backtested	varies	varies	Negative	1–5 days
Carry / yield / vol-selling	C — no single index	No single figure — huge, diverse category	high, but	tail: 30–96%	Strong negative	weeks–months
Buy & hold S&P 500	A — investable index	10.2–10.5% §	0.4–0.5 (full-cycle)	50%+ (2008)	Neutral / neg	forever

Red-shaded cells: no defensible public benchmark exists — these strategies must be individually defined and backtested before any return can be claimed.  All figures full-cycle. Verify each via Section 2a.

What the Category column means — and why it protects you A — Real, investable / professional benchmarks (S&P 500, SG Trend Index). The strongest evidence: actual, auditable, net-of-fee track records. Trust these most. B — Academic research factors (momentum, value). Real, peer-reviewed, and computable from public data (Kenneth French library) — but they are gross of trading costs and often not directly investable as-is. Good evidence, with an asterisk. C — No public benchmark (mean reversion, carry/vol-selling). There is no single index for “mean reversion” or “carry” — they are enormous families of very different strategies. Any single CAGR here would be invented. So this article refuses to print one. If a course shows you a precise CAGR for a strategy in Category C, ask exactly which system produced it, over what period, net of what costs — because the honest answer is “it depends entirely on the implementation.”

† Why trend following shows 5%, not the 10–15% you may have seen elsewhere. This is the SG Trend Index — the return of the largest professional managed-futures funds, net of their steep 2%-management + 20%-performance fees. A do-it-yourself trader running their own system avoids that fee drag — but removing fees is not a free upgrade: it does not hand you the institutional execution, financing, diversification and research those funds have. So treat the 5% as the return of one specific index, not a ceiling and not a promise. Note too that this figure is the historical return of that particular index — not the theoretical maximum of trend following. Academic evidence (AQR, “A Century of Evidence on Trend-Following Investing,” testing 1880–2016) finds time-series momentum was profitable in every decade since 1880 at roughly a 0.4 Sharpe net of costs, and paid off in 8 of the 10 worst drawdowns for a 60/40 portfolio — which is the real point: the value is that it tends to work when everything else is bleeding (2008, 2022).

‡ “Market + a few %” is the honest long-run academic figure for the momentum factor (gross, before costs and taxes) — a real but modest premium, not a reliable 10–15%. Important: the 30–73% “max drawdown” is NOT the loss of a normal long-only momentum fund. The famous >73% loss over ~3 months in 2009 refers to the classic academic long/short (winner-minus-loser) momentum factor, a leveraged research construct — not “a momentum system loses 73%.” The lesson is subtler and more useful: a strategy can have an attractive long-run average and still carry catastrophic path risk. Source: Daniel & Moskowitz, “Momentum Crashes,” J. Financial Economics (2016).

§ S&P 500 figure is Total Return (dividends reinvested), since 1926. Stated explicitly so it is never confused with a price-only S&P 500 series, which is roughly 2 percentage points a year lower. Reinvested dividends account for a large share of the long-run total — without them the number would be materially smaller.

 

So is it even worth the effort? — read this before you conclude “no” An honest reading of the table above is: for the average person, no. The effort, screen-time and stress of active trading are usually not worth it versus simply buying a low-cost index fund and holding — which is exactly why ~90% of active professionals fail to beat the index over 15 years, and why the courses have to exaggerate. If the honest numbers talk you out of chasing “10x the market,” they have done their job. The narrow case for doing it anyway rests on things the bare CAGR hides: fees (the 5% is net of institutional 2-and-20; your own system pays none of it, though it also lacks their infrastructure); timing (trend returns arrive during crashes, a diversification value a single number cannot show); and the fact that your real goal is not the managed-futures 5% but to beat the index on a risk-adjusted basis — shallower, shorter drawdowns — with a system you run yourself. That is achievable and worthwhile. Beating the index many-fold is not, and anyone promising it is selling the rainbow, not the pot of gold.

The single most important caveat: Sharpe is not comparable across skew A trend system at Sharpe 0.6 and a vol-selling system at “Sharpe 1.8” are not “the second is better.” Sharpe rewards steady small wins and is blind to a fat left tail, so it flatters negative-skew strategies (mean reversion, carry, vol-selling) and understates positive-skew ones (trend, momentum). This is a documented, structural point — SocGen’s own data shows the SG CTA Index, the S&P 500 and a 60/40 portfolio can have similar Sharpe ratios despite completely different tail risk (the CTA index has positive skew; the S&P 500 and 60/40 have negative skew). Source: CFM, “A Good Time for Trend Following,” and the SG Prime Services indices page. Bottom line: compare within-family, or use drawdown-based ratios (MAR/Calmar), never raw Sharpe across different skews.

 2a.  Sources & How to Verify Each Figure

Every figure in the master table traces to a named, public source. Links are free to access. Where a cell is an estimate rather than a measured index value, that is stated plainly.

Strategy (category)	Source & how to verify
Buy & hold S&P 500  (A)	Total Return (dividends reinvested), ~10.2% since 1926 (officialdata.org / Shiller data). Reinvested dividends are a large share of the total; the price-only series is ~2%/yr lower. Full-cycle Sharpe ~0.44.
Trend following  (A)	SG Trend Index (Société Générale), inception 2000: ~4.98% CAGR, ~20.6% max drawdown; SG CTA Index Sharpe ~0.61. Much higher in good years (+27% in 2022). Academic backing across a full century: AQR, “A Century of Evidence on Trend-Following Investing” (Hurst, Ooi, Pedersen, 2017) — profitable every decade since 1880, ~0.4 Sharpe net of costs.
Cross-sectional momentum  (B)	Academic factor, computable from the free Kenneth French Data Library. The >73% loss over ~3 months in 2009 is the long/short WML factor, not a long-only fund: Daniel & Moskowitz, “Momentum Crashes,” J. Financial Economics (2016).
Value / fundamental  (B)	Academic factor (HML), free from the Kenneth French Data Library; Fama & French (1993, 2012). Underperformed for most of 2010–2020 — the pain is duration, not depth.
Mean reversion  (C)	No single public index exists — “mean reversion” is a broad family of very different short-horizon strategies. This article deliberately prints no CAGR for it. The general profile (high win rate, negative skew, sharp tail losses) is documented, but any specific return must be produced by a precisely defined, backtested system — net of costs, out-of-sample.
Carry / yield / vol-selling  (C)	No single figure: FX carry, bond carry, commodity carry and volatility-selling are distinct strategies with different risk. The defining, verifiable fact is the catastrophic tail — e.g. Feb 2018 “Volmageddon,” when the XIV note lost ~96% overnight (widely reported). One event does not characterise the whole family; the tail risk does.

Categories: A = real investable / professional benchmark; B = academic research factor (gross of costs); C = no public benchmark, must be individually backtested. The article prints no return figure for Category C on purpose.

3.  What Each Strategy Actually Feels Like to Trade

The statistics describe the numbers; this describes the experience — which is what determines whether you hold on through the bad stretch or capitulate at the worst moment.

Strategy	Where the edge comes from	Characteristic failure mode	Best environment
Trend following	Serial correlation: markets adjust to news gradually, so trends persist (see Moskowitz–Ooi–Pedersen, “Time Series Momentum,” 2012)	Death by a thousand cuts — choppy, range-bound markets bleed you through whipsaw. Long flat periods (2010s).	Sustained directional moves; crises (2008, 2022)
Cross-sectional momentum	Relative strength persists — past winners keep beating past losers	Momentum crash — violent junk-rallies after a market bottom. Lost >73% in 3 months in 2009.	Steady trends with dispersion between winners and losers
Mean reversion	Prices overshoot short-term and snap back to a reference level	Dies in one afternoon — a strong trend runs the position over; one big loss erases many small wins	Range-bound, liquid, low-trend markets
Carry / vol-selling	Collecting a premium others pay to offload risk	Catastrophic tail — 2008 FX carry, Feb 2018 vol (“pennies in front of a steamroller”)	Calm, low-volatility, stable regimes
Value / fundamental	Cheap assets re-rate to fair value over long horizons	Pain by duration — can underperform for a decade (2010–2020)	Post-bubble normalisation; rising-rate regimes
Why the “feel” matters more than the backtest number A system you abandon at the bottom of its drawdown has an effective return of zero. Mean reversion’s high backtest Sharpe looks wonderful — until its negative skew delivers a sudden, large loss precisely when people panic and quit. Trend following is the opposite trap: slow, grinding underperformance (the 2010s were poor for it) that tempts you to “fix” the system by cutting winners early — which historically lowers returns and deepens drawdowns. Match the strategy’s pain profile to what you can genuinely tolerate, not to the prettiest statistic on a slide.

4.  The Closest Thing to a Free Lunch: Diversification

These strategies are negatively correlated in their worst moments. Trend following’s best years are often mean-reversion’s and carry’s worst — which is why combining low-correlation strategies smooths the ride more than optimising any single one. It is often called the one “free lunch” in finance — but it is not literally free: in a real crisis correlations can all jump to 1, strategies crowd, diversification caps your upside, and two “different” strategies can secretly share the same hidden risk. Treat it as the closest thing to free, not a guarantee.

Period	Trend following	Mean reversion	Carry / vol-selling	Equity momentum
2008 crisis	Strong (+)	Hurt	Catastrophic (−)	Mixed
2009 rebound	Weak / negative	Strong (+)	Recovering	Crash (−), −73% in 3 mo
2010–2019 grind	Poor / flat	Good	Good (until 2018)	Mixed
Feb 2018 vol spike	Positive	Hurt	Catastrophic (−)	Hurt
2022 selloff	Strong (+), SG Trend +27%	Hurt	Hurt	Weak

SG Trend Index +27.3% in 2022 per SocGen / AlphaWeek 2022 review; momentum 2009 crash per Daniel & Moskowitz (2016).

Takeaways for a new trader

•    Distrust any promise of “multiples of the market.” The largest professional trend funds have compounded at ~5% a year since 2000, net of fees. Beating the S&P by a small, consistent, risk-adjusted margin is a genuine achievement; beating it many-fold is marketing.

•    Judge against the right benchmark. The S&P 500 does ~10% a year with 50%+ drawdowns. “Good” means a similar return with shallower, shorter drawdowns — measured by MAR/Calmar, net of costs, out-of-sample.

•    Respect skew and the tail. A high Sharpe on a negative-skew strategy is hiding its risk where Sharpe can’t see it. The blow-up is not “if” but “when.”

•    Verify everything. Every number here has a public source. If a course won’t show you audited, out-of-sample, net-of-cost results with drawdowns, assume the results don’t exist.

 

5.  The $10,000 Experiment: What “Doubling Your Money” Really Implies

The fastest way to see through an advertisement is compounding arithmetic. Every return has a doubling time (roughly 72 ÷ the annual %). Modest returns double your money slowly; the advertised numbers imply impossibilities.

Annual return	Approx. doubling time	$10,000 after 10 years	$10,000 after 20 years
5%	~14.2 years	$16,300	$26,500
7%	~10.2 years	$19,700	$38,700
10% (≈ S&P)	~7.3 years	$25,900	$67,300
15%	~5.0 years	$40,500	$163,700
20%	~3.8 years	$61,900	$383,400
25%	~3.1 years	$93,100	$867,400
50%	~1.7 years	$576,600	$33.3 million
100% (“double every year”)	1 year	$10.2 million	$10.5 billion
Why this kills the “100% a year” pitch instantly If anyone could truly double their money every year, $10,000 would become $10.5 billion in 20 years. There would be a dozen new richest people on earth every generation, all self-taught traders. There aren’t. The arithmetic itself proves the claim is impossible as a sustained rate — not merely unlikely. Even a “modest-sounding” 25% a year, held for 20 years, turns $10,000 into over $860,000; sustained by almost no one alive. Warren Buffett, one of the greatest investors in history, compounded at roughly 20% over decades — and that made him one of the richest people on the planet. When an ad promises more than Buffett, the ad is the product, not the returns.

Rule of thumb to verify the table yourself: doubling time ≈ 72 ÷ annual return %. Future values use (1 + r)^years.

6.  The Backtest & Survivorship Trap: 10 Questions for Any “Track Record”

Suppose someone shows you: “My system made 35% a year for 15 years.” Before believing it, ask these ten questions. Most advertised results quietly fail several of them — and each failure inflates the number.

Ask…	Why it matters
1.  Was it live money, or only a backtest?	A backtest is a hypothesis, not a track record. Real money reveals slippage, fills, and discipline a simulation hides.
2.  Was the strategy changed during the period?	If rules were tweaked as it went, you’re seeing hindsight, not an out-of-sample result.
3.  Were losing versions quietly discarded?	Test 100 systems, show the one that worked, and you’ve “discovered” noise. This is survivorship bias in strategy selection.
4.  Were delisted / bankrupt stocks included?	Testing only companies that still exist deletes the losers — and massively overstates returns.
5.  Were realistic commissions and fees included?	High-turnover systems can look brilliant until costs are deducted, then break even or worse.
6.  Was slippage included?	The price you backtest at is rarely the price you get, especially in size or in fast markets.
7.  Was the strategy optimised on the same data it’s shown on?	Curve-fitting to history guarantees a beautiful backtest and tells you nothing about the future.
8.  Was any of it genuinely out-of-sample?	A result that holds on data the system never saw is worth far more than one that doesn’t.
9.  Was the result produced before or after the strategy was chosen?	A number cherry-picked after the fact is not evidence; it’s selection.
10.  What was the worst drawdown, and how long to recover?	A 35% average with a 60% drawdown is unholdable for most people — the average is a fiction you never actually earn if you quit.
The human factor — the number the backtest never shows A system you abandon at the bottom of its drawdown has an effective return of zero. The deepest problem with advertised backtests usually isn’t fraud — it’s that the mathematical system and the human being have different results. A backtest that earns 30% a year might run: +42%, +18%, −27%, −19%, +8%… The maths made money. But most people quit in year four, at the bottom, and lock in the loss. The backtest has no fear, no mortgage, no spouse asking why the account is down 40%. You do. That is why “can I actually hold this through its worst stretch?” matters more than the headline CAGR — and why a lower-return system you can stick with beats a higher-return one you’ll abandon.

This is not just a story — it is measured. Morningstar’s annual “Mind the Gap” study tracks the difference between what funds returned and what their investors actually earned. Over the 10 years to end-2024, the average fund returned 8.2% a year but the average investor earned only 7.0% — a persistent ~1.2-percentage-point-a-year “investor return gap,” about 15% of the total return, lost to the timing of buying and selling. (In fairness, researchers debate how much of that gap is truly bad timing versus normal cash-flow patterns — one study puts the pure-timing cost far lower — but the direction is consistent: real investors, trading real emotions, tend to trail the very funds they hold.)

Source: Morningstar, “Mind the Gap 2025” (10 years to Dec 2024).  If a professionally-run fund’s own investors lose ~1.2%/yr to their own behaviour, a leveraged retail system in a beginner’s hands will fare worse, not better.

How the ads fake the number: leverage Most “50–100% a year” pitches are not a better strategy — they are an ordinary strategy run at 10× to 50× leverage. This is the mechanism behind the impossible arithmetic in Section 5. Leverage multiplies the return distribution — both tails — without improving the underlying edge: the Sharpe ratio, win rate and skew stay the same, but every gain and every loss is scaled up. A strategy earning 8% at a 0.7 Sharpe becomes “40% a year” at 5× — and simultaneously acquires a real path to −100% (total ruin) from a drawdown that would have been a survivable −16% unleveraged. So when an ad shows a huge CAGR, the right question is not “what’s the edge?” but “how much leverage, and what is the risk of ruin?” High advertised returns usually signal high leverage — which means the marketing is selling risk, dressed up as skill.

The litmus test for any strategy claim

If an advertised system claims a CAGR above ~25% over a multi-year period, ask to see three things:

•    Audited live-money track records — not a backtest.

•    The maximum drawdown, and how long it took to recover.

•    The transaction-cost and slippage assumptions behind the numbers.

If they cannot — or will not — provide all three, the claim is marketing, not trading data. Walk away.

Sources & further reading

•    SPIVA U.S. Year-End 2025 Scorecard — S&P Dow Jones Indices  (active vs. passive, free, newest edition)

•    SPIVA U.S. Persistence Scorecard — S&P DJI  (do winning funds keep winning? mostly not)

•    SG Prime Services CTA / Trend Indices — Société Générale  (managed-futures benchmarks)

•    AQR — “A Century of Evidence on Trend-Following Investing” (Hurst, Ooi, Pedersen, 2017)  (trend following 1880–2016)

•    Kenneth R. French Data Library — Dartmouth  (raw factor returns for momentum, value, size)

•    S&P 500 total return since 1926 — officialdata.org (Shiller data)

•    Daniel & Moskowitz, “Momentum Crashes,” Journal of Financial Economics 2016

•    Morningstar — “Mind the Gap 2025” (investor return gap)  (why investors trail their own funds)

•    Morningstar Active/Passive Barometer; Fama & French (2010), “Luck versus Skill in the Cross-Section of Mutual Fund Returns.”

Disclaimer: This article is for education, not investment advice. Index figures are historical, depend on start date, currency and dividend treatment, move over time, and do not guarantee future results. Where no defensible public benchmark exists (mean reversion, carry), this article prints no return figure rather than an estimate. Verify all figures via the linked primary sources before relying on them.