Edge Decay: Why Strategies Stop Working (And When to Notice)

A strategy that worked beautifully for 18 months starts producing flat results. Six months later it's slightly negative. The trader is confused, nothing changed in the strategy. But something changed in the market: the edge decayed. As more participants found the same setup, the alpha got arbitraged away. Real edges don't last forever. Recognizing decay early is what lets you stop trading dead strategies and find new ones.

Why edges decay

The mechanism is simple: edges exist because of inefficiencies. Inefficiencies attract participants who profit from them. As more participants do the same trade, the inefficiency shrinks. Eventually, the inefficiency is gone, and the edge with it.

This is fundamental to how markets evolve. Markets become more efficient over time as participants compete for returns. The edges that worked in 2018 are mostly gone in 2026 because thousands of participants have spent years arbitraging them away.

Edges that decay slowly:

• Behavioral edges (human nature changes slowly)
• Edges requiring specific expertise or infrastructure
• Edges in genuinely small markets that can't absorb large participants

Edges that decay fast:

• Technical patterns that get widely shared
• Narrative-driven trades that become obvious
• On-chain signals that aggregator platforms expose
• Anything that gets popular on twitter

The half-life of an edge varies enormously based on what it is and how visible it becomes. Some edges last decades; others last weeks.

The signs of edge decay

Several signals to watch:

1. Declining expectancy. The clearest signal. Your strategy's average R per trade is shrinking over time. If your rolling 50-trade expectancy was +0.3R two years ago and is +0.05R now, the edge is decaying.

2. Smaller average winner. The strategy still wins at the same rate, but the wins are smaller than they used to be. The market is absorbing the move faster, front-running by other participants is leaving you with smaller R-multiples.

3. Larger or more frequent losses. The setup still triggers, but follow-through is weaker. Losses cluster in patterns that didn't happen before.

4. The setup is "obvious" to others. You see the same setup discussed on twitter, in educational content, in newsletters. The edge that existed when only a few traders knew it doesn't exist when thousands trade it.

5. Strategy's prerequisites change. The market structure that produced the edge has shifted. Spot dominance gave way to derivatives. The on-chain landscape changed with new platforms. Whatever produced the original edge isn't there anymore.

If multiple of these are present, the edge is likely decaying. Time to investigate and decide whether to retire, refactor, or wait it out.

Distinguishing decay from regime mismatch

A strategy underperforming in a wrong regime looks similar to a strategy whose edge has decayed. The distinction matters because:

• Regime mismatch: the edge is intact, just inactive. When the regime returns, performance returns.
• Decay: the edge is gone. The regime returning won't bring it back.

How to tell:

Regime mismatch signals:

• Strategy's underperformance correlates with a clear regime shift
• Other regime-matched strategies are also underperforming
• Historical performance in similar regime conditions was also weak

Decay signals:

• Underperformance persists across multiple regimes
• The setup is now widely shared/discussed
• Even when the prior favorable regime returns, performance doesn't recover
• The mechanism that produced edge no longer exists

Regime mismatch → pause and wait. Decay → retire and move on.

The discipline: don't conflate them. Pausing during regime mismatch is correct (the strategy still works); pausing during decay is incorrect (the strategy is dead and you should redeploy capital elsewhere).

How to respond to decay

Several strategic responses:

1. Retire the strategy. The cleanest response. Stop trading it. Reallocate the attention/capital to other strategies or to developing new ones.

2. Refactor the strategy. Sometimes the underlying insight is real but the specific implementation has decayed. Maybe the new version uses different filters, different timeframes, different assets. Be cautious, refactoring can drift into curve-fitting.

3. Move down market. The edge might still exist on smaller, less-watched assets. The major-pair version is gone, but the small- cap version still has the inefficiency. Trade-off: illiquid markets are harder to execute in.

4. Find a complementary edge. The decayed edge might have been the trigger, the setup that got you positioned. The actual edge might be your filtering, sizing, exit discipline. If those remain, find new triggers and apply the same discipline.

5. Wait and watch. Some apparent decay is actually regime mismatch in disguise. If you're not sure which, reduce size substantially and watch over a longer time horizon. Resume normal sizing only when the strategy clearly performs again.

What you should not do: keep trading the dead strategy at full size while hoping it works again. The data is telling you to act; ignoring the data is the path to larger drawdowns.

A common mistake: treating decay as a temporary slump

A trader's strategy stops working. Their internal narrative: "It'll come back; the market is just weird right now." Six months later, the strategy is still flat or negative. They've lost time and capital trading a dead edge.

The fix: distinguish slump (low-probability random underperformance) from decay (systematic edge loss). Slumps fix themselves within reasonable time; decay doesn't. Set a time limit on how long you'll wait before retiring (e.g., "if expectancy hasn't recovered in 60 trades, the strategy is retired").

A common mistake: refusing to retire validated strategies

Some traders identify so deeply with their strategy that retiring it feels like personal failure. They keep trading it for emotional reasons even after the data shows decay.

The fix: strategies are tools, not identities. The trader who profitably traded a strategy for two years isn't diminished by retiring it when it stops working. The skill that made the strategy work, the rigor, the discipline, the validation, transfers to the next strategy.

A common mistake: chasing the latest "hot" edge

A trader's strategy decays. They look for a new edge. Twitter is full of "AI tokens are the new edge" / "on-chain X is the alpha." They pile into whatever's hot.

The fix: the edges shouted about on twitter are typically already crowded. Real new edges are usually quiet, discovered through your own analysis, not through social media virality. Be cautious about the edges everyone's talking about.

A common mistake: assuming all your edges decayed

A trader has multiple strategies. One decays. They assume all their edges are gone and stop trading entirely. But the others might still be working, they just need to be evaluated separately.

The fix: evaluate each strategy independently. Per-strategy expectancy tracking is what tells you which are working and which aren't. Don't make portfolio-level decisions on single-strategy data.

The meta-skill: detecting decay reliably

The hardest part of edge decay isn't the strategy it's recognizing the decay early enough to act. Several disciplines help:

Continuous expectancy monitoring. Rolling 30-trade expectancy plotted over time. The trend is what matters; a downward trend is the early warning.

Regime tagging. Each trade tagged with the regime it occurred in. This lets you separate decay (decline across regimes) from regime mismatch (decline only in specific regimes).

Comparison to strategy's expected variance. If your strategy was supposed to have a worst case of -X% based on backtesting, and live is approaching -1.5X%, you're outside the validated envelope.

Periodic strategy reviews. Quarterly or semi-annually, formally evaluate each strategy against its expected performance. Decay often shows up gradually; the regular review catches it.

These disciplines are what make decay-detection operational. Without them, you'll notice decay months or years too late.

Mental model, edges as fishing spots that get crowded

Imagine you discovered a great fishing spot. For the first season, you caught fish constantly. By the second season, a few other fishers found your spot. By the third season, the spot was packed. By the fourth, the fish were mostly caught and the catch rate fell to average.

The spot didn't change, the competitive landscape did. You can keep fishing the spot at average rates, or you can find a new spot. Smart fishers spend some time searching for new spots even when the current one is working, so they have backups when the current one gets crowded.

Edge decay is the same dynamic. Spots get crowded. Finding new spots is part of the job, not a side project. Continuously developing potential edges (even when the current one is working) is what keeps you profitable across decay cycles.

Why this matters for trading

Edge decay is the multi-year reality of trading. Every strategy you trade now will eventually stop working, the question is when, not if. Hex37's journal data gives you the per-strategy expectancy tracking that makes decay detection possible. The discipline of "track expectancy by strategy, retire when decay clearly persists, develop new strategies continuously" is what differentiates traders who profit across multiple cycles from traders who profit in one cycle and never adapt.

Takeaway

Real edges decay over time as more participants find them. Detect decay via declining expectancy, smaller winners, larger losses, "obvious" setup discussion, changed market structure. Distinguish decay (edge gone) from regime mismatch (edge inactive). Respond by retiring, refactoring, moving down market, finding complementary edges, or waiting (only if regime mismatch). Set time limits on how long you'll wait before retiring. Continuously develop new edges so you have backups when current ones decay. The edges that lasted decades are rare; most last 1-5 years. Plan accordingly.