A fully systematic strategy has an advantage that discretionary traders often envy: every decision follows written rules, so you can explain exactly why it trades. That clarity is valuable, but it isn't proof. When an allocator looks at an algorithm, the question isn't "is it systematic?" It's "do the live results behave the way the rules say they should, and is there enough of them to judge?"
This guide looks at a rules-based strategy through an allocator's eyes, then sets out the routes open to algorithmic and quant traders who want to manage outside capital.
Algorithmic traders access investor capital by turning a backtested idea into a live, verified track record that investors and allocators can assess. Allocators look for three things: live results that match the strategy's stated logic, stable execution that holds up as size grows, and a record long enough to cover many independent decisions and more than one kind of market. The main routes are allocation programmes, running a fund or managed accounts within the rules, and platforms where a verified strategy can be offered to investors, such as a DARWIN.
Rules remove one source of doubt: the trader's mood on the day. They add others. A rules-based strategy can be overfitted to past data, built on optimistic cost assumptions, or tuned until a backtest looks smooth. Nobody outside can tell which from the description alone.
An investor can't see your code, and usually wouldn't want to. What they can see is the live record: the equity curve, the trades, the drawdowns and the risk taken along the way. So an allocator reads a systematic strategy as a set of claims, and the live record either supports those claims or contradicts them.
The checks below apply to almost any systematic approach, from a simple trend follower to a multi-model portfolio.
|
What they check |
The question behind it |
What good evidence looks like |
|---|---|---|
|
Live results match the logic |
Is the edge the one you describe? |
A trend system earns in trends and gives some back in ranges. Win rate, holding time and trade frequency sit close to the design. |
|
Stable execution |
Does the strategy survive real fills? |
Slippage and costs in line with your assumptions. No clusters of missed, late or duplicated trades. |
|
Consistent risk |
Will tomorrow's risk look like yesterday's? |
Position sizing that follows the rules. Drawdowns within the range the design implies. No sudden jumps in leverage. |
|
Enough history |
Is this skill or luck? |
Many independent decisions, across more than one market regime. |
|
Capacity |
Can it take more money without changing? |
Results that hold as order size rises. Trading in liquid instruments and hours. |
None of these is about the headline return. A modest return with all five in place usually interests an allocator more than a spectacular curve with gaps in it.
Imagine two traders who deploy the same hypothetical breakout system on a major currency pair, with identical rules and the same risk per trade.
Over time, Trader A's live record looks like the backtest. Holding times, win rate and the shape of the drawdowns are close to the design, and the losses cluster in ranging markets, as a breakout system's should.
Trader B's record has gaps. The trades skipped around announcements were some of the system's biggest winners in testing, so the record shows fewer trades and a different return profile. Trader B's total return might even look similar. But the record no longer describes the system. It describes the system plus Trader B's interventions, and an allocator has no way to model those.
For an algorithmic trader, execution isn't a technical detail. It's part of what an investor buys. The areas that most often separate a backtest from a live record are:
In our investable attributes, Capacity estimates how much additional investment a DARWIN can absorb before divergence starts to hurt investors' returns. It tends to favour longer holding times, wider stops and targets, and orders fed into the market gradually rather than in one block. For algo traders who want to work on this, we've published an open-source order splitter for MetaTrader that slices large orders into smaller, time-spaced pieces.
There is no single number of months. What counts is the number of independent decisions and the variety of conditions they were made in.
A system that trades several times a day builds statistical evidence faster than a swing system that trades a few times a month. But speed isn't everything: both need to show how they behave when volatility, trend strength and liquidity change. A useful test is whether your record includes at least one stretch the strategy wasn't designed for, and whether the drawdown it produced stayed within the range you'd expect.
Our Experience attribute reflects this idea: it grows with the number of representative decisions a DARWIN has made. It's also why resetting an account after a bad run is so costly. A restarted record throws away evidence, and allocators notice.
Once the evidence is in place, there are four broad ways to put it in front of capital. Each asks something different of you.
There are two ways to build a DARWIN with us:
One distinction matters. Our DarwinIA programme makes notional allocations to DARWINs based on a monthly ranking. That is separate from investor capital, which comes from external investors who choose to invest in a DARWIN. Neither is guaranteed, and past performance doesn't guarantee future results. See the current DarwinIA rules for how it works today.
Which platforms are best for algorithmic traders who want to access capital? Look for a platform that captures your record from live or virtual trading data, lets you keep and build on that record over time, and gives investors a fair way to compare strategies. Also check that it supports the tools you trade with and doesn't restrict the way your system works.
Can an expert advisor (EA) build a verified track record? Yes. A track record belongs to the account, whether trades are placed by code or by hand. What matters is that the record is captured from real trade data and that the EA's live behaviour matches the logic you describe.
Do investors care whether a strategy is automated? Less than you might think. Quant and algo traders often attract investors because their process is repeatable, but investors judge the live record first: consistency, risk, drawdowns and how long it has run.
Is a good backtest enough to attract investors? Rarely. A backtest shows what a strategy might have done. Investors want evidence of what it did with real execution, real costs and no hindsight.
What's the difference between a DarwinIA allocation and investor capital? DarwinIA allocations are notional allocations we make to DARWINs based on a monthly ranking. Investor capital comes from external investors who choose to invest in a DARWIN. They're separate, and neither is guaranteed.
If you want to build a verified record for your system without putting your own capital at risk, start a Darwinex Zero membership and run it on virtual capital. You can see the current plans on the Darwinex Zero pricing page.