Testing / Sample report
Sample report · illustrative

What the report you receive looks like

Below is a full report, just as a client receives it: the verdict, the charts, all the indicators and the recommendations. Every figure is explained in plain language — you don’t need to know trading terms to see what works, what doesn’t and what to do next.

Illustrative example. The idea and the figures are made up, on simulated data, so as not to disclose anyone’s work. They are not the results of a real strategy, of Maonta or of a client.

1The verdict, in short

Verdict

Promising, with conditions

ViablePromising, with conditionsDoesn’t pass

The idea has a real edge, but smaller than it looked on the data it was built on, and it is sensitive to costs. With two adjustments (see the recommendations) it has a chance to become viable. As it stands, we wouldn’t put real money on it without a real-time test.

+4.2%average return per year, on the unseen period (2023–2026)
−18.1%the largest drop of the account
0.75return-to-risk ratio (Sharpe)
+2.3%return per year if costs were double
89%of random strategies did worse
9 of 11years in profit, over the whole history
6.3out of 10

Overall score

Promising, with conditions

Not yet: it is worth taking further with the proposed adjustments, but it is not a strategy to put real money on right now.

Return4,2+4.2% a year on the unseen period; the market made +6.2%
Stability7,09 years out of 10 positive; return/risk ratio (Sharpe) 0.75
Drawdowns6,3largest drawdown −18.1%; at most 282 trading days below the peak
Real edge7,989% of random strategies did worse
Robust rules9,116 out of 16 nearby variants of the rules ended positive
Cost resilience5,6with double costs the return drops from +4.2% to +2.3% a year
Consistency4,8+8.7% a year on the build period, +4.2% on the unseen one

Strengths

  • Consistent results: almost every year ended positive.
  • The result is not luck: it beats the vast majority of random strategies.
  • The rules are not fragile: any nearby variant gives similar results.

Weaknesses

  • The return is modest and below the market: the idea has to add something over buy and hold.
  • Sensitive to costs: more expensive execution takes more than half of the return.
  • On unseen data it made about half of what it showed while being built: part of the result was fitting to the past.

How we calculate the score: each criterion gets 0–10 by fixed rules, the same for any strategy; the overall score is their average, in which return, drawdowns and real edge count double. The score says how well the idea passed the tests, not how much it will earn.

2The idea and the rules

This is how the client described the idea, in their own words:

“When a big stock drops sharply in one day, I buy it the next day and hold it for a few days, because it usually recovers.”

Together, we turned it into exact rules, so it can be tested without interpretation:

Which stocks
the 100 largest US companies, using each month’s list (not today’s)
The signal
the stock fell by at least 4% in one day
The entry
buy at the next day’s open
The exit
sell at the close of the third day
Position size
20% of the account per stock, at most 5 positions at once
Costs
0.17% per round-trip trade, measured on real fills
Build period
2016–2022: where the idea was designed and checked
Unseen period
2023–2026: kept untouched and used only once, at the end

The rules were sealed before any calculation, with the digital fingerprint sha256: 7c41…e09a. Any later change would have changed the fingerprint — that’s why the result can’t be “tweaked” after we see it.

3The account over time

What would have happened to 10,000 € put into the strategy at the start of 2016, compared with the same amount simply held in the market (the large-stock index).

the strategythe market, simulated (just buy and hold) the unseen period

What you see here

The strategy reached 21,380 €, the market 27,978 €. The market made more, but went through much bigger drops: −37.8%, against −18.1% for the strategy.

What it means for you

The strategy is a calmer road, not a faster one. On the unseen period it made +4.2% a year, against +8.7% on the build period: the edge halved. That’s normal — and it’s the figure to expect in reality.

4How low the account went

The chart shows, at every moment, how far the account was below the highest level reached up to then. Zero means a new record.

the strategythe market

What you see here

The strategy’s largest drop was −18.1%, and the account stayed below its peak for at most 282 trading days (about 13 months). The market fell as far as −37.8%.

What it means for you

Ask yourself honestly: if you see 8,192 € out of 10,000 € and it lasts for months, would you stick with the strategy? If not, risk less per position (see section 10).

5Year by year and month by month

the strategythe market 2026: up to the end of September

The result of every month, in percent (green: profit, red: loss):

What you see here

9 years out of 11 ended in profit, and 58% of the months. The best year was 2021 (+22.6%), the weakest 2026 (−6.1%).

What it means for you

Profit doesn’t come evenly: there are weak months and even weak years. A good strategy isn’t one that never loses, but one that makes money in most years and has no losses big enough to knock you out of the game.

6The trades

Trade results on the unseen period

how many trades ended with each result, in percent, after costs
613trades in total, about 55 a year
57%ended in profit (unseen period)
+4.6%average gain of a winning trade
−5.2%average loss of a losing trade
1.17profit factor: gains divided by losses
+0.39%expected gain per trade, after costs

The 5 largest losses on the unseen period, each in just 3 days:

  • 2025-12-03: −21,7%
  • 2025-09-05: −15,0%
  • 2024-11-08: −14,3%
  • 2025-12-01: −12,3%
  • 2026-08-18: −12,0%

What you see here

More than half the trades win, but the average loss is slightly larger than the average gain. Out of every 100 euros gained, losses take back about 85.

What it means for you

A few very large losses (over 10% in 3 days) spoil the result. This is where there’s the most room for improvement — see recommendation 4.

7Luck or edge? The control test

We ran 1,000 “strategies” that enter at random, just as often, with the same costs and the same holding time. If the idea had no edge at all, its result would look like theirs.

the 1,000 random strategies the idea being tested

What you see here

The idea made +4.2% a year on the unseen period. The random strategies made, in the middle, −1.4% a year. The idea did better than 89% of them.

What it means for you

It’s a good sign, but not yet proof: our threshold for “viable” is beating at least 95% of the random strategies. The difference may come from a real edge, but also from a bit of luck.

8The costs

Every trade costs money: the difference between the buying and selling price, the commission, the execution. We measured them on real fills, then checked what remains if they were double or triple.

return per year on the unseen period, at different cost levels

What you see here

With no costs, the idea would make +6.0% a year. With the real costs, +4.2%. At double costs, +2.3%, and at triple almost nothing: +0.5%.

What it means for you

The strategy passes the double-cost test, but without much margin. It matters which broker you use and how you place orders: an expensive broker or “market” orders on low-volume stocks can eat the whole profit.

9How fragile the rules are

We changed the rules slightly — the drop threshold and the number of days — to see whether the result holds only at the chosen values or around them too. Return per year on the unseen period:

The drop that triggers entryhold 2 dayshold 3 dayshold 4 dayshold 5 days
−3%+3,3%+3,0%+3,3%+2,1%
−4%+3,3%+4,2%+3,9%+3,2%
−5%+3,8%+3,8%+3,6%+2,7%
−6%+2,5%+3,8%+2,1%+2,7%

What you see here

All 16 variants ended in profit, between +2.1% and +4.2% a year. The chosen rule (−4%, 3 days, framed) is among the best, but its neighbours aren’t much weaker.

What it means for you

A very good sign: the idea isn’t “fitted exactly” to the past. If only one combination had worked and the rest hadn’t, it would almost certainly have been luck.

10How much to risk

The same strategy, depending on how much of the account you put into one position. The return grows, but so do the drops, in almost the same proportion (over the whole history):

Position sizereturn per yearlargest drop worst dayfor whom
10% of the account per position+3,5%−9,4%−1,4%very prudent: small drops, small return
15% of the account per position+5,3%−13,9%−2,0%the balance we recommend
20% of the account per position (the tested rule)+7,1%−18,1%−2,7%as the client proposed
30% of the account per position+10,7%−26,0%−4,1%aggressive: a drop of a quarter of the account

What you see here

With 15% per position instead of 20%, the return falls from +7.1% to +5.3% a year, but the largest drop shrinks from −18.1% to −13.9%.

What it means for you

Position size doesn’t make a strategy better — it only shifts the balance between return and peace of mind. Choose the level at which a drop like the one in the table wouldn’t make you give up.

11All the indicators, in plain language

Each indicator, on the build period, on the unseen period (the one that matters) and, for comparison, for the market on the same unseen period.

Indicatorbuild
2016–2022
unseen
2023–2026
market
(unseen)
What it meansHow it rates
Average return per yearannual ROI · CAGR+8,7%+4,2%+6,2%How much the account grew, on average, each year, with gains reinvested.acceptable
Total returncumulative return+82,6%+17,1%+26,4%How much the account grew over the whole period. Over the whole history, 10,000 € would have become 21,380 €.acceptable
Largest dropmaximum drawdown−10,7%−18,1%−26,9%How far the account fell from its highest point before recovering. It’s the figure you’ll feel the most.watch
Longest time below the peakdrawdown duration209 days282 days579 daysHow many trading days the account stayed below its previous record. 282 days means about 13 months of patience.watch
Return-to-risk ratioSharpe1,630,750,46How much return each unit of account swings brings. Below 0.5 is weak, around 1 is good, above 1.5 is very good.acceptable
Return-to-drops ratioSortino1,800,770,63Like Sharpe, but counting only the downward swings — the ones that hurt.acceptable
Return versus the largest dropCalmar0,810,230,23The return per year divided by the largest drop. Above 0.5 is good; below 0.3 means big drops for the return obtained.weak
Yearly swingsvolatility5,2%5,6%16,0%How much the account moves in an ordinary year. Lower means a calmer road.good
Winning tradeswin rate56%57%—Out of 100 trades, how many ended in profit. It matters together with the size of the gains and losses.good
Average gain / average losswin/loss ratio+4,6% / −4,0%+4,6% / −5,2%—How much a winning trade brings on average, and how much a losing one takes.watch
Gains divided by lossesprofit factor1,421,17—The sum of all gains divided by the sum of all losses. Above 1 means the strategy makes money; above 1.3 is comfortable.borderline
Expected gain per tradeexpectancy+0,75%+0,39%—How much a trade brings on average, after costs. It must be clearly above the cost of one trade.borderline
Trades per yearfrequency5752—How many trades the strategy makes in a year. More trades = more costs.good
Time in the marketexposure52%49%100%On how many days the strategy had at least one open position. The rest of the time, the money sits quietly.good
Worst daylargest daily loss−1,7%−2,7%−7,4%How much the account lost on its worst day.good
Worst monthlargest monthly loss−6,6%−6,0%−7,6%How much the account lost in its worst month.good
Months in profitconsistency61%53%60%Out of 100 months, how many ended in profit.acceptable

12Our recommendations

In order of importance. The first two don’t change the idea; the last two are adjustments that must be tested separately, with the rules sealed again.

Expect +4.2% a year, not +8.7%

The unseen period is the closest to reality. Plan your money and your patience around it, not around the nice figures of the build period.

Keep costs at the tested level

With limit orders and a low-commission broker, you stay at +4.2% a year. At double costs the return falls to +2.3%, and at triple it disappears.

Risk 15% per position, not 20%

You give up a little return, but the largest drop shrinks to −13.9%. For most people, that’s the difference between sticking with the strategy and quitting at the worst moment.

Test a loss limit per trade

The largest losses were between −12.0% and −21.7% in just 3 days. An exit at −8% could cut these tails — or it could take the strategy out of trades that recover. Only the test will tell.

proposed adjustment · tested separately

Test a market filter

Don’t buy when the whole market is falling (below its 200-day average). In such periods, big stock drops are slower to recover.

proposed adjustment · tested separately

What comes next

The client chose to test both adjustments (recommendations 4 and 5), then to run the best version for 6 weeks on a virtual account, day by day. Only after that is it worth talking about real money — and only a little at first.

That’s what your report looks like: the same sections, the same explanations, but with your idea and your figures.

Illustrative example on simulated data: it does not represent the results of a real strategy, of Maonta or of a client. Results of a test on historical data do not guarantee future results. This is not investment advice.