The strategy database

Capture sustained moves across commodities, rates, FX, and equity indices using normalized price breakout signals.

Risk: Drawdowns 15–30% in backtests; diversification across markets essential.

Value, momentum, quality and low volatility take turns failing, and they tend to fail at different times, so blending them into one score produces a far smoother ride than any single factor can.

Risk: Blending can quietly cancel your own bets, since value and momentum often point at opposite stocks. Every factor in the blend is public and crowded, and 2018 to 2020 showed that they can all underperform together. Weighting the factors is where most overfitting hides.

Trade the spread between cointegrated pairs when it deviates from the equilibrium; exit at mean reversion or stop.

Risk: Cointegration breaks; one leg can gap; liquidity in both legs critical.

Outperform by switching between equities and cash (or bonds) when relative and absolute momentum favor the switch.

Risk: Max drawdown ~20% in backtests; regime shift risk.

Target a stable portfolio volatility by scaling exposure to a diversified multi-asset basket inversely to realized vol.

Risk: Leverage increases in low-vol regimes; drawdown when vol spikes.

Earn the interest rate differential by going long high-yield and short low-yield currencies, with vol scaling.

Risk: Carry unwinds in risk-off; correlation with equity stress.

Sell index puts on a fixed schedule and hold cash against them, collecting the premium that buyers pay for downside insurance.

Risk: You are selling crash insurance. Losses are highly non-linear: dozens of small wins, then one month that erases years. Margin can spike exactly when you can least afford it, forcing you out at the worst price. Gap risk means stop losses do not reliably work.

Capture roll yield in backwardated commodity futures by holding front-month (or optimized roll) with vol scaling.

Risk: Contango erodes returns; single-commodity risk; roll timing matters.

Rotate into sectors with strongest relative momentum (e.g. 3–6m) and out of weakest; rebalance monthly.

Risk: Concentration risk in few sectors; underperforms in rapid rotation.

Rank assets across equities, bonds, commodities, FX by 12m momentum; long top quartile, short bottom quartile.

Risk: Correlation in drawdowns; need liquid instruments and execution.

Gross profit divided by total assets predicts future returns about as well as book to price, because it measures productive quality before accountants get a chance to obscure it.

Risk: Profitable companies are usually expensive, so the raw signal fights value and can leave you long crowded quality names at the top of a cycle. The metric is sensitive to how a company classifies costs, and it is close to meaningless for banks and insurers.

Small caps overshoot on short-term moves; fade extreme z-scores over 5–10 days for mean reversion edge.

Risk: Blow-up risk in strong trends; use strict stop and size limits.

Investors systematically underpay for safe, profitable, growing, well-run companies, so buying quality and shorting junk earns a premium that shows up across markets and decades.

Risk: Quality is expensive quality most of the time, so the strategy can turn into a crowded bet on a handful of compounders. It lags badly in junk rallies off market bottoms. The definition of quality is highly discretionary, which invites overfitting.

Cheap stocks beat expensive ones, but only if you compare a bank to a bank; ranking value within sectors removes the accidental sector bet that ruins most naive value screens.

Risk: Value can underperform for a decade at a stretch, as it did through the 2010s. Cheap often means genuinely broken, so value books fill up with value traps. Accounting-based signals are stale by construction and are distorted by intangibles-heavy business models.

Buy a market if its own past 12-month return was positive and sell it if negative; slow-moving investor behaviour and risk transfer make past direction weakly predictive of future direction.

Risk: Long stretches of flat performance when markets chop. Sharp losses at trend reversals because the position is largest right before the turn. Signal is simple and widely known, so returns have thinned versus the 1980s and 1990s.

Buy options when implied volatility is cheap relative to expected realized, then hedge the delta repeatedly and harvest the difference from the market's own wiggles.

Risk: This is the long-volatility side, so the tail is in your favour, but the bleed is relentless. You pay theta every single day. If the market goes quiet you lose steadily and predictably, and most traders quit right before the move that would have paid for everything. Hedging costs can silently consume the entire theoretical edge.

Spend a small, fixed slice of the portfolio every year on far out of the money puts so that a crash makes you money instead of ending you.

Risk: This strategy loses money in most years, by design. The bleed is constant and demoralising, and the single hardest risk is behavioural: investors abandon the hedge after a few flat years, usually shortly before it would have paid. Monetisation timing is genuinely difficult, and a hedge you never sell is a hedge that expires worthless.

Instead of one stock against one stock, hedge a target name against a weighted basket of peers and trade the leftover spread when it stretches.

Risk: A basket hedge is only as good as the weights. Weights estimated on old data drift, and a single broken leg (M&A, fraud, delisting) turns a hedged position into a naked one. Crowding in classic peer baskets means everyone exits the same spread at once. Borrow cost and recall on the short legs can quietly eat the whole edge.

Exploit pricing inefficiencies between prediction markets and crypto derivatives across exchanges.

Risk: Latency-sensitive; execution and one-leg fill risk; capital fragmentation; regulatory divergence. Historical data limited.

Stocks that beat earnings expectations keep drifting up for weeks afterwards, and stocks that miss keep drifting down, so buy the surprises and hold them through the drift.

Risk: The effect has weakened materially as more capital chases it, and much of the remaining drift sits in small, illiquid names where costs eat the edge. Earnings dates and consensus data are riddled with look-ahead traps. Crowding means many participants are positioned identically into the same reports.

Screen for names where TTM EPS is rising quarter-over-quarter and current P/E is below its historical median, then investigate why mispricing may exist.

Risk: EPS quality and regime shifts can invalidate median P/E; cheap can persist. Use risk controls.

Follow trend in rates (e.g. 10Y yield or bond futures) with vol-scaled position; capture sustained rate moves.

Risk: Central bank regime shifts cause whipsaws; rate volatility spikes.

Every futures curve has a slope, and that slope is a forecast of the return you earn if prices do not move; go long the markets with the most favourable slope and short the least favourable, across all asset classes at once.

Risk: Carry is a crowded trade that unwinds violently in risk-off periods. Positive carry frequently compensates for a real risk that eventually shows up. Correlated with equity drawdowns despite looking diversified.

Strip out the market, sector and factor moves from each stock's recent return, then fade only what is left, because the leftover part is the piece most likely to be temporary pressure.

Risk: The whole strategy depends on the risk model being right. If a real factor is missing from the model, its returns leak into the residual and you will systematically fade a genuine trend, which is a slow and expensive way to lose. Turnover is high, so costs dominate. Crowded among quant market-neutral funds, which means shared drawdowns during deleveraging events.

Implied variance is priced above the variance that actually gets realized, so selling variance and holding to expiry earns the spread.

Risk: Variance swaps pay on squared returns, so a single large day dominates everything. A short variance position has losses that grow with the square of the move: a 10 percent drop hurts four times as much as a 5 percent drop. Payoffs are unbounded on the downside. Caps are essential and still may not save you.

Investors fixate on reported earnings without checking whether real cash came in; companies whose profits are mostly accounting entries go on to underperform the ones whose profits are backed by cash.

Risk: The signal has decayed substantially since it was published in 1996 and much of what remains sits in small, hard-to-trade names. The short leg holds aggressive-accounting companies that can keep running for years before they break, and shorting them is dangerous and expensive.

The same company listed in two countries should be worth the same amount once you convert the currency; when the two prices drift apart by more than the cost of converting between them, trade the gap.

Risk: The gap is usually small and often reflects a real barrier: capital controls, conversion restrictions, taxes or borrow scarcity. Currency risk must be hedged or it dwarfs the trade. In emerging markets the deviation can persist for months, and the reason it persists is exactly the reason you cannot arbitrage it. Non-overlapping trading hours mean apparent gaps are frequently stale prices, not opportunities.

Carry earns steadily and crashes in shocks; trend loses slowly and profits in shocks; running both in one futures book produces a far smoother return than either alone because their bad periods rarely overlap.

Risk: The two signals can point the same way and double a position, which concentrates risk exactly when both are confident. Blending does not remove the possibility that both lose together, which happened in choppy inflationary periods. Combination weights are easy to overfit.

Currencies that have strengthened over the past several months tend to keep strengthening, so rank the majors on past return and go long the leaders against the laggards.

Risk: Momentum reverses violently at turning points, and central bank intervention can flip a trend overnight. The strategy is quietly correlated with carry because high yielders often also have the strongest trends. Crowding is heavy in G10 FX, and the edge has visibly decayed since the 2008 era of pinned rates.

When one company has two classes of stock with nearly identical economics, the price gap between them should be stable, so trade it when it stretches beyond its normal range.

Risk: There is no forced convergence date. The gap can widen for years and there is nothing that makes it close, which is what makes this trade genuinely dangerous despite looking safe. Voting rights, index membership and liquidity differences are real reasons for a persistent gap, not errors. Borrow on the expensive class is often the binding constraint, and the classes may never be equalised.

Collect the large interest rate premium in emerging market currencies while paying a small, constant insurance premium against the crash that eventually comes.

Risk: Carry goes up by the stairs and down by the elevator. Emerging currencies can lose 20 to 40 percent in weeks. Capital controls, NDF fixing disputes, sudden devaluations, illiquid crisis markets and counterparty risk on forwards are all live risks. The hedge cost is a permanent drag that will make quiet years look boring.

An ETF should trade at the value of the stocks it holds; when the price and the underlying basket drift apart, trade the gap and let the creation and redemption process close it.

Risk: The clean version of this trade belongs to authorised participants with a cost base you cannot match. Everyone else is trading a residual that is usually smaller than fees, spreads and borrow. Discounts can widen and stay wide when the underlying market is closed, halted or illiquid, and that is exactly when the trade looks most attractive.

Go long the government bond markets that pay you the most to wait and are cheapest on a real yield basis, short the ones that pay least, and stay neutral to the global level of rates.

Risk: Every developed bond market can fall together, so a market-neutral construction is essential and still incomplete. Central banks intervene directly in these markets through quantitative easing and yield curve control, which can hold a market away from fair value for years. Currency exposure must be hedged or it will dominate the return.

Index options are usually expensive relative to the options on the stocks inside the index, so sell index volatility and buy the components.

Risk: Short implied correlation is short crash risk. In a market-wide selloff every stock moves together, correlation goes to one, and the index leg loses far more than the single-name legs gain. Execution across dozens of chains is expensive and operationally heavy. Not accessible without institutional infrastructure.

Instead of betting on one fast and slow moving average pair, run many pairs at once and average their signals, which keeps the trend edge while removing the luck of a single parameter choice.

Risk: Averaging reduces parameter luck but does not remove trend risk itself. Faster pairs add turnover and cost. In sideways markets the blended signal still hovers near zero and bleeds.

Compare each company only against its direct competitors, buy the cheap ones and short the rich ones inside the same industry, so the industry itself cancels out and only the relative call remains.

Risk: Cheap relative to peers often means deservedly cheap, and the strategy will repeatedly buy structurally declining businesses. Convergence can take years and there is no catalyst forcing it. Crowded among market-neutral equity funds, and the short leg carries borrow cost and squeeze risk. Value spreads can widen for very long periods.

After a takeover is announced the target usually trades below the offer price, so buy the target and collect that gap as the deal closes, accepting the risk that some deals collapse.

Risk: The payoff is small and steady until a deal breaks, and then the loss is large and immediate. Target shares typically fall back to, or below, their pre-announcement price. Regulatory blocks, financing failure and hostile political intervention have all killed deals that looked certain. Deal breaks cluster in market selloffs, so the strategy loses most when everything else does.

Boring stocks quietly beat exciting ones on a risk-adjusted basis, so a long-only portfolio built to minimise total volatility earns close to market returns with far less pain.

Risk: Structurally overweights defensive sectors and behaves like a bond proxy, so it gets hurt when rates rise sharply. Badly crowded after a decade of ETF inflows. The optimiser is highly sensitive to covariance estimation error, which can produce concentrated, fragile portfolios.

Let the data itself define the factors, hedge each stock against those data-driven factors, and trade the leftover residual when it drifts too far from its own average.

Risk: Statistical factors are unstable: the components you extract this month may not be the ones that exist next month, and they have no economic meaning to fall back on when they break. Choosing how many components to keep is an unavoidable judgement call that quietly determines everything. Heavily crowded, and shares its worst days with every other quant market-neutral book.

Rank stocks on the part of their past return that factor exposures cannot explain, so you buy real winners instead of accidental bets on beta, size or sector.

Risk: Still a momentum strategy, so it can crash hard when a beaten-down market snaps back. Turnover is high, borrow on the short leg can be expensive or unavailable, and the regression window choice quietly drives most of the results.

Hold the index and sell calls against it on a fixed schedule, converting some upside into steady option premium.

Risk: This is a short-volatility, capped-upside trade. You keep all of the downside of the index and give away the best up months. Premium looks steady until a fast rally truncates your gains or a crash hits the equity leg. Options liquidity and bid-ask spread drive a lot of the realized edge.

VIX futures usually trade above spot VIX, so a short position in the front futures earns carry as the price rolls down toward spot.

Risk: Short VIX futures is one of the most dangerous positions available to a retail trader. VIX can more than double in a single session. February 2018 destroyed short-vol products and their holders in one afternoon. Leverage and ETP rebalancing mechanics can force liquidation at the worst tick.

When short-dated implied vol is expensive relative to long-dated, sell the front expiry and buy the back, and let the term structure normalise.

Risk: The short front leg carries most of the gamma, so a large move before front expiry can produce a loss far bigger than the spread cost suggests. Inverted term structure can stay inverted or invert further. Vega on the long back leg means a collapse in overall vol hurts you even if the slope goes your way.

Most of the improvement in a managed futures programme comes from how positions are sized and risk-budgeted, not from the trend signal, so build the portfolio layer deliberately rather than treating it as an afterthought.

Risk: Volatility scaling reacts to volatility, not to direction, so it will not save you from a correlated reversal. Low-volatility periods produce high leverage, and the de-levering after a shock always happens too late.

Sell volatility (e.g. covered calls or put spreads) when implied vol is rich vs realized; manage tail with defined risk.

Risk: Short vol carries tail risk; black-swan events can wipe out premium.

Buy short-dated bonds and sell long-dated ones when the curve is unusually flat or inverted, betting that the spread between two and ten year yields widens back out.

Risk: The curve can stay inverted for two years while you bleed negative carry every single day. A steepener put on too early is the classic way to lose money while being fundamentally right. The position is duration-neutral only on paper: in a real selloff the long end moves more than the model expects.

Investors who cannot borrow money buy risky stocks instead, which overprices high-beta names; going long low-beta stocks with leverage and short high-beta stocks harvests that distortion.

Risk: The strategy uses leverage on the long leg, so a funding squeeze or margin call can force liquidation at the worst moment. It loses badly in sharp junk rallies and in deleveraging events. Beta estimates are noisy and the strategy is only as good as the estimate.

Bucket stocks by lowest price rank and test returns under realistic liquidity + circuit constraints.

Risk: Microcaps and penny stocks: liquidity, circuits, governance and pump-and-dump risk. Backtest must model execution constraints.

Trade one contract month against another in the same commodity to bet on the shape of the curve rather than the direction of the price, isolating the storage and scarcity signal while cancelling most of the flat-price risk.

Risk: Spreads look calm until they are not. Margin offsets encourage heavy leverage, and a squeeze in the front month can move a spread by many times its usual daily range. Liquidity in deferred months is far thinner than it appears.

When a dated crypto future trades well above spot, buy the coin, short the future, and lock in the gap as a known return that converges to zero at expiry.

Risk: The locked-in return is only locked in if both legs survive to expiry. Exchange insolvency, withdrawal freezes and margin calls on the short leg during a rally are the real killers. On offshore venues the counterparty is often unregulated. Basis can widen further before it converges, forcing losses on a mark-to-market basis even though the final payoff is fixed.

When crypto perpetual futures trade above spot, longs pay shorts a funding fee every few hours, so hold spot and short the perpetual to collect that fee with no directional exposure.

Risk: This is not a risk-free yield. Exchange failure, withdrawal freezes and outright fraud have wiped out participants in this exact trade. Funding can flip negative and you pay instead of collect. A sharp rally can liquidate the short leg if collateral is not managed. Stablecoin depegs, oracle failures and auto-deleveraging are all live and have all happened.

Apply simple momentum (e.g. 20d/55d MA cross or 12m trend) to BTC and ETH with strict vol cap and size limits.

Risk: Extreme vol and drawdowns; exchange and custody risk; regulatory uncertainty.

Implied vol on single names runs up into earnings and collapses the morning after, so sell the event premium and let the crush pay you.

Risk: Gap risk in its purest form. A single stock can move 25 percent overnight on a guidance change, and a naked short straddle in that name loses far more than every winning trade combined. Single-name option spreads are wide. Assignment risk is real. This is a strategy where one bad name can end the whole book.

A standard commodity index mechanically holds the front contract and rolls on a fixed calendar, which is a known and exploitable pattern; choosing a better point on the curve and a better roll date captures most of the index return with far less roll cost.

Risk: Still fully long commodity beta, so it falls with commodity prices. The enhancement is a cost saving, not a hedge. Deferred contracts are less liquid, and the roll timing edge shrinks as more money uses it.

Run a short-lookback trend system across liquid futures purely as a crisis hedge; it loses small amounts most of the time and pays out during extended market declines, because crises are slow enough for a fast trend system to turn short in time.

Risk: Expected to lose money in most calendar years. Very sensitive to whipsaws and to transaction costs. Provides no protection against a one-day crash, only against multi-week declines. Requires the patience to hold a losing sleeve for years.

Stocks trading near their 52-week high keep outperforming, because the round-number high acts as a psychological ceiling that makes investors slow to accept good news.

Risk: A close cousin of price momentum, so it inherits the crash risk. The signal is anchored to a single historical price point, which makes it fragile around splits, spin-offs and thin trading. Crowded, cheap to compute, and therefore widely arbitraged.

Currencies that have drifted far from what local prices justify tend to drift back over multi-year horizons, so buy the cheap ones and sell the expensive ones and wait.

Risk: Value can stay wrong for five years or more, which is longer than most investors or most careers can wait. The signal fights carry directly, because cheap currencies are usually cheap for painful reasons. Structural breaks such as a permanent productivity shift or a commodity discovery can make an apparently cheap currency correctly cheap forever.

A stock index future should equal the index plus financing minus dividends; when the future strays outside that band, buy the cheap side, sell the rich side and hold to expiry.

Risk: The theoretical mispricing is almost always inside the cost band for anyone who is not a large bank. Real risks are financing, dividend forecast error and margin calls before expiry, none of which show up in the textbook formula. Persistent basis richness usually reflects balance sheet scarcity at banks, which is a cost you also face rather than a gift you can collect.

When a stock is added to a major index, passive funds must buy it on a known date, so position ahead of that forced demand and exit into it.

Risk: This is one of the most crowded and best-known event trades in equities, and the pre-announcement drift has largely been arbitraged away. Buying a stock that will be added is a directional position with no fundamental support, and if the addition is cancelled or the flow is smaller than expected, the unwind is ugly. Front-running has become so competitive that the effect frequently reverses after the effective date.

Stocks that fell hardest over the last week tend to bounce and stocks that jumped hardest tend to give some back, because much of the move was liquidity demand rather than news.

Risk: This is a liquidity-provision strategy dressed up as an alpha signal, so it loses exactly when liquidity is scarce and it is most tempting. Turnover is very high and transaction costs consume most of the gross return. Buying the worst performers means repeatedly catching names that are falling for real reasons: fraud, dilution, going-concern doubt.

Downside puts are chronically more expensive than equidistant upside calls, so sell the rich put and buy the cheap call to harvest the skew.

Risk: A risk reversal is short the crash and long the melt-up. You are being paid to take exactly the risk that shows up in every equity bear market. The put you sold gains value fastest exactly when your other assets are falling. Skew can also steepen against you long before any crash arrives, producing mark to market losses with no crash at all.

Quiet markets store energy; when price breaks out of a tight range by more than a multiple of recent daily movement, the move often continues, so enter in the direction of the break rather than fading it.

Risk: High false-breakout rate. Win rate is often under 40 percent and the strategy relies on a few large winners. Stops get run in thin markets and around scheduled news. Costs are heavy because turnover is high.

Certain dates in the calendar have historically shown persistent return patterns, such as the turn of the month and the days around holidays, and a systematic overlay can tilt exposure toward them.

Risk: This is the field of quantitative finance most contaminated by data mining. Most published calendar effects do not survive out of sample, and several famous ones vanished the moment they were published. Effects are small relative to costs, and any single year can go entirely the wrong way. Treat every result here as guilty until proven innocent.

Within commodities alone, buy the futures that have outperformed their peers over recent months and short the laggards; supply and demand imbalances take months to resolve, so relative strength within the complex persists.

Risk: Small universe means limited diversification, roughly 20 to 30 markets and only a handful of independent sectors. Reversals are sharp. Several markets are thin and expensive to trade. Weather and geopolitics can invalidate a signal overnight.

The gap between nominal and inflation-linked bond yields is the market's inflation forecast, and it systematically over- or under-shoots, so trade the gap rather than the level of rates.

Risk: Breakevens contain a liquidity premium that collapses in a crisis, so the trade loses hard in exactly the moments a macro hedge is supposed to help. Inflation-linked bonds are far less liquid than nominals and can gap. The position is duration-neutral in theory and rarely so in practice. March 2020 saw breakevens collapse for liquidity reasons that had nothing to do with inflation.

The average CTA return is largely explained by a handful of simple trend signals on liquid futures, so a transparent rules-based portfolio can track the industry index at a fraction of the fees.

Risk: Replication tracks the average manager, including in bad years. Regression weights are unstable and fit the recent past, so the tracker is always one regime behind. It cannot capture whatever any individual manager does well.

The raw size premium is mostly an illusion created by tiny unprofitable companies dragging returns down; screen the junk out of small caps and a genuine premium appears.

Risk: Small caps are illiquid, expensive to trade and brutal in a crisis, when they fall further and recover later than large caps. The unscreened size premium has been close to zero for decades, so the whole strategy depends on the quality screen actually working.