Performance

What the strategy would have returned, and what it cost to hold.

The published rules, applied to market data from July 2010: what they produced, what holding them required, and what they cost.

These are simulated results. No money was managed over this period, and nothing here is a record of executed trades. The four qualifications that bound every figure on this page are set out directly below the record.

The record

What the simulation produced

Annualised return

19.0%

Compounded after costs, over the whole period. It is what the curve compounds to, not a figure any account received.

July 2010 to September 2026. One series, one convention: $10,000 at the start, every change applied on the published rules, five basis points of cost per side. The curve is on a logarithmic axis, so the same distance always means the same multiple.

How did it grow over time?Growth of $10,000, simulated

Simulated, 2010-07-13 to 2026-09-21. Logarithmic axis, weekly sample.

Show the figures (one point per year, of 847 weekly)
DateValue of $10,000
2010-07-13$10,000
2011-07-06$13,685
2012-07-04$16,441
2013-07-03$23,092
2014-07-02$27,282
2015-07-01$28,247
2016-06-29$32,104
2017-06-28$37,595
2018-06-27$44,483
2019-06-26$53,087
2020-06-24$62,850
2021-06-23$84,687
2022-06-22$89,807
2023-06-21$99,623
2024-06-19$120,999
2025-06-18$138,847
2026-06-17$162,758
2026-09-21$167,055
Deepest fall
-9.8%
From a high, before recovery
Volatility
8.4%
Annualised

Read this first

Simulated, not traded

The rules are applied to historical prices. No account followed them over this period, and no order was placed. Results a real account would have obtained differ.

Presented on a reduced basis

The figures shown apply a deliberate reduction to the periodic returns of the model, and the whole curve is rebuilt from them. Volatility and drawdown fall with the return, because they are properties of the same path. This is the only series this page shows.

Costs included

Five basis points of transaction cost per side, applied on every change. Taxes, spreads and the cost of your own broker are not included.

Not the author's own use

Nicolas has followed a portfolio personally since May 2026. That is a private use, on a short period, and it is not part of this series and not audited.

Year by year

Not every year looks like the average

An annualised figure is an average, and no single year is the average. Over the whole period the simulation has no losing year, and the first and last are partial: they are kept, because removing them would flatter that count.

Which years gained, and which lost?Return by calendar year

2010 and 2026 are partial years.

Show every year
PeriodValue
20107.1%
201137.1%
201218.9%
201352.9%
20142.9%
20153.4%
201618.2%
201731.1%
20184.4%
201915.1%
202041.9%
202121.9%
20221.7%
202315.3%
202419.6%
202519.8%
20268.4%
Show every quarter (65)
QuarterReturn
2010Q3Partial period4.5%
2010Q42.5%
2011Q10.8%
2011Q226.2%
2011Q32.4%
2011Q45.3%
2012Q111.0%
2012Q20.3%
2012Q37.9%
2012Q4-1.0%
2013Q128.9%
2013Q22.2%
2013Q33.7%
2013Q411.9%
2014Q1-0.2%
2014Q22.1%
2014Q3-1.1%
2014Q42.1%
2015Q11.2%
2015Q21.5%
2015Q3-2.2%
2015Q42.9%
2016Q17.6%
2016Q25.4%
2016Q31.1%
2016Q43.1%
2017Q15.3%
2017Q25.7%
2017Q36.9%
2017Q410.2%
2018Q1-0.4%
2018Q21.8%
2018Q34.0%
2018Q4-1.0%
2019Q17.5%
2019Q26.4%
2019Q30.1%
2019Q40.5%
2020Q16.1%
2020Q212.8%
2020Q35.8%
2020Q412.1%
2021Q17.0%
2021Q26.1%
2021Q30.0%
2021Q47.3%
2022Q1-3.4%
2022Q21.0%
2022Q3-1.3%
2022Q45.6%
2023Q14.1%
2023Q23.3%
2023Q3-1.5%
2023Q48.9%
2024Q18.7%
2024Q23.2%
2024Q33.6%
2024Q42.8%
2025Q10.5%
2025Q29.5%
2025Q36.1%
2025Q42.6%
2026Q10.0%
2026Q24.8%
2026Q3Partial period3.4%

The cost of holding

How far it fell below its own high

A return you never received is easy to accept. A fall you watched is not. This is the distance below the previous high, month by month, for sixteen years. The weekly figures are in the table.

How far below its own high did it fall?Distance below the previous high

Monthly. The trough shown is the deepest fall of the whole period, unchanged by the sampling.

Show the figures (one point per month)
DateBelow previous high
2010-07-130.0%
2010-07-29-0.4%
2010-08-26-0.1%
2010-09-27-0.1%
2010-10-21-0.9%
2010-11-24-0.0%
2010-12-16-0.9%
2011-01-18-2.9%
2011-02-11-2.9%
2011-03-16-2.2%
2011-04-12-0.9%
2011-05-05-1.7%
2011-06-02-0.9%
2011-07-01-1.5%
2011-08-01-0.1%
2011-08-26-1.6%
2011-09-22-3.0%
2011-10-20-1.4%
2011-11-23-3.6%
2011-12-19-3.9%
2012-01-13-0.8%
2012-02-10-0.5%
2012-03-08-0.3%
2012-04-10-2.7%
2012-05-09-2.7%
2012-06-03-6.1%
2012-06-28-1.9%
2012-07-26-0.6%
2012-08-23-2.5%
2012-09-26-1.5%
2012-10-24-2.9%
2012-11-15-5.2%
2012-12-15-2.5%
2013-01-10-1.1%
2013-02-070.0%
2013-03-13-0.5%
2013-04-040.0%
2013-05-02-7.9%
2013-06-05-8.3%
2013-06-27-9.8%
2013-07-30-7.0%
2013-08-22-7.1%
2013-09-24-6.2%
2013-10-17-5.0%
2013-11-20-1.8%
2013-12-12-2.0%
2014-01-13-1.0%
2014-02-08-2.3%
2014-03-08-0.6%
2014-04-07-1.5%
2014-05-07-1.5%
2014-06-02-0.9%
2014-06-26-0.1%
2014-07-30-0.5%
2014-08-21-0.5%
2014-09-23-1.1%
2014-10-16-2.8%
2014-11-13-0.4%
2014-12-16-1.7%
2015-01-14-0.8%
2015-02-09-0.1%
2015-03-11-1.5%
2015-04-02-1.2%
2015-04-30-0.7%
2015-05-29-0.6%
2015-06-25-1.1%
2015-07-27-2.4%
2015-08-25-4.8%
2015-09-23-4.0%
2015-10-21-2.5%
2015-11-13-4.7%
2015-12-10-3.9%
2016-01-12-4.3%
2016-02-09-2.7%
2016-03-08-0.7%
2016-04-05-1.0%
2016-05-04-1.3%
2016-05-31-0.1%
2016-06-27-3.3%
2016-07-25-0.7%
2016-08-24-0.4%
2016-09-16-1.4%
2016-10-17-2.5%
2016-11-14-1.1%
2016-12-14-0.7%
2017-01-11-2.0%
2017-02-02-0.8%
2017-03-08-1.7%
2017-04-05-0.7%
2017-05-03-0.1%
2017-05-27-1.2%
2017-06-27-1.2%
2017-07-24-0.0%
2017-08-21-1.3%
2017-09-20-0.1%
2017-10-17-0.3%
2017-11-12-2.3%
2017-12-09-0.6%
2018-01-10-0.1%
2018-02-05-5.1%
2018-03-01-4.4%
2018-04-02-6.3%
2018-05-02-5.1%
2018-05-29-3.8%
2018-06-27-3.8%
2018-07-19-1.7%
2018-08-16-1.4%
2018-09-17-1.0%
2018-10-11-4.2%
2018-11-14-2.6%
2018-12-06-2.3%
2019-01-03-2.3%
2019-02-06-0.1%
2019-03-06-1.0%
2019-03-28-0.6%
2019-04-25-0.8%
2019-05-23-0.9%
2019-06-20-0.3%
2019-07-23-0.4%
2019-08-15-0.3%
2019-09-18-0.3%
2019-10-15-0.9%
2019-11-11-1.3%
2019-12-09-1.3%
2020-01-08-0.2%
2020-02-04-0.5%
2020-02-28-1.2%
2020-04-01-3.8%
2020-04-23-1.7%
2020-05-24-0.8%
2020-06-24-3.6%
2020-07-16-0.3%
2020-08-19-0.6%
2020-09-10-3.7%
2020-10-08-2.4%
2020-11-07-1.0%
2020-12-09-0.8%
2021-01-04-0.8%
2021-01-31-2.2%
2021-02-28-3.7%
2021-03-25-1.6%
2021-04-22-0.7%
2021-05-21-0.3%
2021-06-18-2.1%
2021-07-19-1.8%
2021-08-18-1.4%
2021-09-10-1.4%
2021-10-12-2.7%
2021-11-10-0.7%
2021-12-03-2.9%
2022-01-05-1.1%
2022-01-27-5.5%
2022-02-25-3.2%
2022-03-28-3.2%
2022-04-27-3.8%
2022-05-19-4.5%
2022-06-21-2.1%
2022-07-14-3.3%
2022-08-17-0.9%
2022-09-14-2.7%
2022-10-12-6.9%
2022-11-03-3.9%
2022-12-07-1.9%
2023-01-04-1.1%
2023-01-30-0.4%
2023-03-01-1.5%
2023-03-28-0.6%
2023-04-26-1.5%
2023-05-24-1.0%
2023-06-21-0.5%
2023-07-16-0.5%
2023-08-16-2.1%
2023-09-13-1.5%
2023-10-05-4.0%
2023-11-02-2.2%
2023-12-06-0.5%
2024-01-03-1.2%
2024-01-31-0.9%
2024-02-220.0%
2024-03-26-0.4%
2024-04-19-2.7%
2024-05-22-0.2%
2024-06-130.0%
2024-07-17-0.9%
2024-08-08-3.8%
2024-09-06-2.7%
2024-10-07-0.7%
2024-11-03-1.3%
2024-11-28-0.7%
2024-12-31-2.0%
2025-01-27-1.1%
2025-02-25-2.0%
2025-03-21-4.0%
2025-04-21-3.1%
2025-05-21-0.6%
2025-06-18-0.7%
2025-07-15-0.4%
2025-08-07-0.6%
2025-09-05-0.1%
2025-10-07-0.1%
2025-11-04-1.4%
2025-11-27-0.5%
2025-12-31-0.9%
2026-01-22-0.5%
2026-02-23-0.9%
2026-03-20-2.0%
2026-04-21-0.5%
2026-05-19-1.9%
2026-06-17-2.9%
2026-07-13-3.6%
2026-08-06-1.3%
2026-09-09-1.3%
2026-09-21-0.8%

Time below a previous high

82.2%

This is the figure most readers do not expect. Being below a previous high is the ordinary state of a rising curve, not the exception: the fall is shallow, and it lasts. A holder who judges the strategy by whether it is at a new high will conclude it is failing most of the time.

The three that asked the most

Any fall of more than five per cent from a high, until that high is regained. The rule was set before the series was read, which is why the deepest of them is also the deepest fall of the whole period, already shown above.

01-9.8%June 2013

Began April 2013. Back to its previous high after 211 days.

02-6.9%October 2022

Began August 2022. Back to its previous high after 108 days.

03-6.7%February 2018

Began January 2018. Back to its previous high after 210 days.

Show every episode (9)
StartedDeepestFallDays to recoverRecovered
April 2012June 2012-6.1%90Yes
October 2012November 2012-5.2%104Yes
April 2013June 2013-9.8%211Yes
January 2018February 2018-6.7%210Yes
September 2020September 2020-6.1%71Yes
January 2022January 2022-5.5%205Yes
August 2022October 2022-6.9%108Yes
July 2024August 2024-5.4%64Yes
February 2025March 2025-5.6%77Yes

Whenever you had started

The result barely depends on the day you began

A calendar year starts on one date. A rolling window starts on every date. Counting all of them is the only way to answer the question a reader actually has, which is not how the strategy did in 2017 but how it did if they had been the one holding it.

Positive 12-month periods

99.9%

Of every rolling twelve-month window in the period, measured on the daily series. One window in that count is negative, and its figure is beside it.

Worst twelve months
-0.8%
Cumulative over the window
Longer windows

Each figure is the total return over the whole window, not a yearly rate. The annualised equivalent is in its own column.

WindowWindows countedWeakest, cumulativeMedian, cumulativeStrongest, cumulativeMedian, annualised
1 year5,550-0.8%17.3%59.7%17.3%
3 years4,81919.7%63.5%154.0%17.8%
5 years4,08968.9%125.1%186.9%17.6%

What it asks of you

How often the allocation is traded

Every change costs something, and a reader who has to place the orders is entitled to know how many there are before subscribing. These figures are not a measure of quality: a strategy that traded twice as often would not be twice as good.

Weeks with a trade
25
Out of 52, on average
Weeks with nothing to do
52.0%
Of all publication cycles
Weight turnover
5.3
Sum of the weights traded, each year

Method

What is in the number, and what is not

What the simulation applies

  • The published rules, unchanged over the whole period.
  • Prices from the reference market data, adjusted for splits and dividends.
  • Five basis points of transaction cost per side, on every change.
  • Any share not allocated is held in cash.

What it does not include

  • Taxes, which depend on where you live and on your account.
  • The spread and the fees of your own broker.
  • The difference between a published price and the price you obtain.
  • Any judgement applied after the fact to the rules themselves.

The allocation is published every Wednesday

Members receive the target allocation and the note that explains it. Subscriptions are not open yet.

Witan Way is an independent, rules-based investing education platform. Nothing published constitutes personalised investment advice, a solicitation, or a recommendation to buy or sell. Capital is at risk. © Witan Way.