Ask a trader how diversified their portfolio is, and the answer usually comes back as a list of symbols: gold, a few forex pairs, an index or two. Different markets feel like diversification, and to some extent they are. But market choice is only one lever among several, and it is often the least reliable one on its own. Two strategies trading completely different instruments can still behave almost identically if their underlying logic is similar enough.
Markets and asset classes are a starting point, not the whole picture
Trading different symbols does reduce some forms of risk — a strategy on EUR/USD isn’t exposed to a gold-specific supply shock, for instance. Spreading across asset classes (currencies, indices, commodities, and so on) goes further, since different asset classes can respond differently to the same macro event. This is genuinely useful. It is just not sufficient on its own, because it says nothing about how each strategy actually trades.
Two strategies on two different symbols can both be short-term trend followers that enter after a breakout and exit on a fixed profit target. If broad market momentum picks up at the same time across correlated instruments, both strategies can enter, run, and exit in a similar rhythm — producing return streams that move together far more than the different tickers would suggest.
The other sources of variety that matter just as much
Genuine behavioural diversification tends to come from a wider set of dimensions:
- Direction. A portfolio built entirely from long-biased strategies is exposed to one direction of risk, however many symbols it covers. See Long, Short and Timeframe Diversification.
- Timeframe. A strategy holding trades for minutes behaves very differently from one holding trades for weeks, even on the same instrument — different noise sensitivity, different exposure windows, different reaction speed to news.
- Entry behaviour. Breakout entries, pullback entries, and mean-reversion entries respond to different market conditions, sometimes in opposite ways at the same moment.
- Exit behaviour. Fixed targets, trailing stops, and time-based exits change how a strategy’s trades are distributed through time, and how quickly it reacts to a move reversing.
- Holding period and trade timing. Strategies that tend to be active at different times of day, or hold positions over different lengths of time, are less likely to have all their trades open at once.
- Market regime sensitivity. Some logic works best in trending conditions, some in ranges, some in high volatility — knowing which is which tells you more than the symbol does.
- Strategy logic itself. Ultimately, the underlying rules — what triggers an entry, what defines risk, what defines an exit — are the real source of behavioural difference, and symbol/asset class are downstream of that.
A simple example
As an illustrative example: imagine two EAs, one trading EUR/USD and one trading GBP/USD, both entering on the same type of 15-minute breakout pattern with similar risk management. Despite the different symbols, they are likely to be highly related in practice, because EUR/USD and GBP/USD often move together and the entry logic is functionally the same pattern applied twice. Now imagine a third EA on EUR/USD that trades a slow, multi-day mean-reversion approach. Despite sharing a symbol with the first EA, it may behave far more differently — because the logic, timeframe, and market conditions it depends on are genuinely distinct.
The lesson isn’t that symbol choice is meaningless — it isn’t — but that it is one input among several, and on its own it can give a false sense of variety.
Why this matters for portfolio construction
If diversification is judged only by counting symbols or asset classes, a portfolio can look spread out on paper while still carrying a concentrated behavioural bet underneath. That concentration tends to surface at the worst possible time — during a market condition that happens to challenge the shared logic across several “different” strategies simultaneously. Correlation in Algorithmic Portfolios looks at one practical way of checking for this after the fact, using how strategies’ historical returns have actually moved relative to one another — not just where they trade.
Where this fits in the SIPS workflow
Portfolio Analysis reports a portfolio’s diversification across several dimensions at once — asset mix, market mix, timeframe mix, direction mix, and behaviour mix — precisely so that a strategy library can be assessed on more than symbol variety. Strategy Metrics Explained covers the individual measures that feed into that picture.
The practical takeaway
Before treating a portfolio as diversified, it’s worth asking a more specific question than “how many markets does it cover?” Ask how the strategies actually enter, how long they hold, which direction they lean, and what conditions they depend on. Symbol variety is a reasonable first filter. It is not the finish line.