Skip to content
← Back to home

Equity lab

A working step from portfolio analysis

An example from our daily work: this is how we look at an equity curve internally, change capital and risk and compare two periods. With several set files we lay the same view on top of each other to see which run best together.

Simulated performance (backtest). Past and simulated results are not a reliable indicator of future results. No investment advice, no offer.

Settings

Capital model

Final balance

710,569

USD

Net

+610,569

USD

Return p.a.

18.2%

per year, geometric

Max. drawdown

-6.0%

on closed-trade balance

Profit factor

1.85

Trades

14,927

69.6%

Losing years

0

2015–2026

Period

11.74

years

All figures are without trading commission.

Equity curve

0188k377k565k754k201520172019202120232025

Period 1Period 2

The split line can be moved freely and only compares two periods. It makes no statement about when the system was developed or verified.

Return per calendar year

11%157%1612%177%1817%1921%2014%2118%2225%2323%2438%2524%26

Period 1 against period 2

Period 1
Period 2
Period
01/01/2015 – 31/12/2020
01/01/2021 – 28/09/2026
Return p.a.
12.2%
24.8%
Max. drawdown
-6.0%
-5.9%
Profit factor
1.43
2.06
Trades
7,855
7,072
Win rate
67.0%
72.4%

Basis of this example

  • An automated trading system on one trading instrument; instrument, provider and system name are anonymised.
  • Period 1 January 2015 to 28 September 2026 (about 11.7 years), 14,927 closed trades.
  • Test on real tick data (data quality 99.9%), 20 ms execution delay.
  • In the test: deposit USD 100,000, risk level 0.3% per entry zone, profits reinvested.
  • Swap is included. All figures are calculated without trading commission, which differs from broker to broker.

How this is calculated

  • The basis is the recorded individual trades of the test. Each trade is scaled by the chosen risk and applied to the chosen capital.
  • This is an extrapolation, not a new simulation. Lot rounding, minimum lots and simultaneously open positions are not modelled. At the default values the result matches the test.
  • Drawdown is measured on the balance after each closed trade. The tester reports equity-based drawdown separately, and it is higher.
  • Simultaneous positions and losing streaks weigh disproportionately at higher risk.

Simulated performance of an example system. It is not the result of actual trading, contains no real market activity with client money and guarantees nothing for the future. Trading financial instruments involves a significant risk of loss. This page is not investment advice and not an offer.

Several lines together: portfolio view

This is how we check internally which lines run best together: switch lines on and off, change capital and lot size and see how curve, drawdown and losing years of the combination change. Monthly profits of the lines are compared. The lower the correlation, the more the lines offset each other.

All figures are without trading commission.

Lines
Capital model

Final balance

56,028

USD

Net

+46,028

USD

Return p.a.

15.9%

geometric

Max. drawdown

-25.9%

on a daily basis

Profit factor

1.22

Trades

4,582

Losing years

1

2015–2026

Period

11.67

years

Equity curve of the selection

8k21k34k47k60k201520172019202120232025
CombinationLine ALine BLine CLine DLine E

Net per calendar year (USD)

  • 2015+2,091
  • 2016+4,932
  • 2017+1,479
  • 2018+1,303
  • 2019-152
  • 2020+4,915
  • 2021+3,478
  • 2022+3,022
  • 2023+3,525
  • 2024+5,410
  • 2025+8,094
  • 2026+7,931

Correlation of monthly profits

ABCDE
A1.000.860.440.260.38
B0.861.000.410.280.37
C0.440.411.000.520.53
D0.260.280.521.000.63
E0.380.370.530.631.00

Values near 1 mean two lines win and lose at the same time. Values near 0 or below offset each other.

The lines one by one

LineNet (USD)Max. DDProfit factorTradesLosing years
Line A+9,806-5.3%1.662201
Line B+9,541-4.9%1.951600
Line C+5,810-16.6%1.101,1392
Line D+11,804-12.8%1.201,6160
Line E+9,067-10.4%1.131,4471
  • The basis is five recorded tests (one trading instrument, every tick, data quality 99.9%, period from 1 January 2015, spread included). Instrument, platform and line names are anonymised.
  • Profit per trade is taken unchanged from the reports. Lot size is scaled linearly, and with reinvestment it is chosen so that a stop-loss costs the set share of the balance: an extrapolation of the recorded trades, not a new simulation. Lot rounding and margin are not modelled.
  • Drawdown is calculated daily from realised profits. Open positions within a day are not included.