Max Concurrent Positions and Correlated Risk

Max concurrent positions is one of the most common risk settings and one of the easiest to satisfy without reducing risk at all.

What the setting assumes

A cap on simultaneous positions implicitly assumes each one is an independent draw. Under that assumption, five positions at one unit each carries less risk than one position at five units, because the outcomes are uncorrelated and losses do not coincide.

The assumption is doing all the work. When positions are correlated, they move together, and five correlated positions behave like one position five times the size — with the added cost of five spreads crossed instead of one.

The setting counts positions. Risk depends on exposure. Those are the same thing only when correlation is low.

Where correlation hides

It is rarely obvious from a position list, and it arrives through several channels at once.

Index membership. Large-cap US equities move substantially with the index. Positions across several such names share a dominant common factor regardless of the companies involved.

Overlapping instruments. SPX and SPY track the same index. QQQ overlaps heavily with the largest S&P constituents. Positions across them are closer to the same position than to different ones.

Sector and theme. Names within a sector move on shared news, and thematic exposure cuts across sector labels.

Directional bias. The simplest and most overlooked: several long calls on different underlyings are all long the market. Different tickers, one directional bet.

Volatility exposure. Options positions carry exposure to implied volatility as well as direction. A volatility spike affects every option position simultaneously, including ones with opposing directional views. Two positions that look hedged on direction can lose together on volatility.

Strategy correlation. If several positions came from the same strategy, they share whatever conditions that strategy depends on. Running one strategy across five symbols is not diversification; it is the same logic applied five times.

Correlation is not stable

The part that makes this genuinely hard rather than merely subtle.

Assets that behave independently in calm conditions tend to move together in stressed ones. Correlations rise precisely when diversification is supposed to help, because in a sharp sell-off participants reduce exposure broadly rather than selectively.

So a portfolio that looks diversified on historical correlations may not be diversified on the day it matters. Any position limit justified by measured correlation is justified by a number that changes under stress, in the unfavourable direction.

Setting a limit that means something

Count exposure rather than positions.

Group correlated instruments and limit the group. Treat SPX, SPY, and closely tracking products as one bucket with a shared cap, rather than as three independent slots.

Limit directional exposure, not just position count. A cap on net long or short exposure across all positions constrains the thing that actually matters. Five positions all long is a different risk from five positions with mixed direction, and a position counter cannot tell them apart.

Limit per strategy as well as overall. Positions from one strategy share its assumptions, so a per-strategy cap prevents one idea occupying the whole book.

Size the aggregate, not the slot. This is where the divide-by-20 rule does its work: available trading capital divided by twenty as the ceiling on any single position. The discipline is that the divisor should reflect worst-case concurrency, not typical concurrency. If you would ever hold more than twenty units of exposure at once — because you average into positions, or hold more concurrently than planned — the divisor needs to go up. A per-position rule with an unbounded position count is not a limit.

The multiplier nobody counts

Correlated positions do not just lose together. They lose together while each pays its own transaction costs.

Five correlated options positions cross five spreads on entry and five on exit. If those spreads have widened — and on short-dated out-of-the-money strikes they widen materially through the session — the cost of expressing one view five times is considerably higher than expressing it once.

The diversification is imaginary. The friction is real.

Implementation notes

Enforce limits in the order-placing component so they apply to every order regardless of which signal source produced it. A limit implemented per strategy cannot see the aggregate.

Evaluate the limit against actual broker positions rather than internal state, since a system that believes it holds three positions while holding five will happily open a sixth.

And check the limit at submission rather than at signal generation. Between a signal and an order, other positions may have opened. Computing exposure across positions on every signal is the kind of work that belongs off the order path, on worker thread pools, so the check does not delay the order it is gating.

The honest limits

Correlation is estimated from history and changes under stress, so any grouping is approximate and wrong in the direction that hurts.

A position limit does not bound loss on any individual position. It bounds how many ways you can be wrong at once, which is a different and lesser thing.

Grouping by obvious relationships misses non-obvious ones. Correlations appear through supply chains, factor exposures, and shared holders in ways a sector label does not capture.

And no limit protects against a market-wide move. When everything falls together, the only variables that matter are total exposure and position size — which is why sizing sits upstream of every other control, and why the post-PDT margin regime matters here specifically: real-time intraday margin monitoring means correlated positions moving together affect buying power continuously rather than at a daily checkpoint.

Frequently asked questions

Why is a max position limit not enough? Because it counts positions rather than exposure. Correlated positions behave like one larger position, so five slots can be one bet.

How do I know if my positions are correlated? Look for shared index membership, overlapping instruments, sector or theme exposure, common directional bias, shared volatility exposure, and positions originating from the same strategy.

Do correlations change? Yes, and unfavourably — assets that behave independently in calm markets tend to move together under stress.

How should I set the limit? Group correlated instruments and cap the group, limit net directional exposure, cap per strategy as well as overall, and size the aggregate rather than the slot.

How does this affect the divide-by-20 rule? The divisor should reflect worst-case concurrency. If you might hold more than twenty units of exposure at once, the divisor needs to be larger.

Disclaimer: This article is educational content about trading mechanics and software engineering. It is not investment advice, financial advice, tax advice, legal advice, or a recommendation to buy or sell any security, nor a recommendation of any strategy, order type, or configuration. Any instruments, settings, or figures named are used solely to illustrate mechanics. Options trading involves substantial risk of loss and is not suitable for all investors. Please read Characteristics and Risks of Standardized Options before trading options. Automated trading systems carry additional risks including software defects, connectivity failures, broker API changes, and outages that may prevent orders from being placed, modified, or cancelled. Stop orders do not guarantee an execution price and stop-limit orders may not execute at all. Past performance does not indicate future results, and no configuration, order type, position-sizing rule, or risk setting can guarantee a profit or prevent a loss.

StaxInvesting LLC sells self-hosted trading software. It is not a broker-dealer, investment adviser, or financial institution, and it does not manage accounts, hold member funds, place trades on behalf of members, or access member brokerage accounts. Members run the software in their own cloud environment, connect their own brokerage accounts under their own credentials, and are solely responsible for their configuration, their credential security, and every trade executed in their account. Third-party platform and broker details described here reflect publicly available information as of publication and are subject to change without notice; always verify against current official documentation. Consult a qualified financial adviser and tax professional regarding your individual circumstances.