Author
Three fair-value estimators for an adjacent crude future, each too weak alone: spread EMA, returns beta, cross-book volume. Weight each by one over its error variance and the shared signal survives while the noise cancels.
The clean linear factor model fails because its weights are not constants. Momentum and carry only pay in the right vol regime, and interactions are the one thing a linear model cannot represent.
Bar research assumes the clock drives the market. It doesn't, events do. Fixed rebalance times are front-runnable and bars hide execution seasonality. Screen on bars, validate on event-driven sims, trade on events.
Every moving average is a low-pass filter, and its lag is the price of stripping noise from a causal signal. Read indicators as filter weights, not chart lines, and stop fooling yourself about them.
Commodities are priced in dollars, so a weaker dollar makes them cheaper abroad, lifts foreign demand, and raises their price, which lifts the commodity currencies. Use it as a directional gate, and count the dollar once.
A cross has no dollar to lean on, so trade it off the relative performance of the two countries' stock indices. Divide them in logs, signal off the ratio, and confirm the equities actually lead the cross.
FX correlation is a regime that feeds on itself until it snaps without warning. Measure it rolling, watch USD pairs move as a bloc, and use "USDCAD is the truth" only when oil is asleep.
Your currency's driver rotates by regime, so a fixed watchlist goes stale. Granger causality asks which candidate actually improves the forecast right now, run multivariate so the dollar doesn't fool you.
Monetary policy is the #1 FX driver, but it moves currencies through expectations, not the rate level. The market forward-discounts the path, so you trade the surprise: a cut smaller than priced rallies the currency.
A currency is pulled by four global drivers and three domestic ones at once, and the market keeps switching which set steers. Read the regime before the news, or trade the right analysis in the wrong direction.
PPP and the Big Mac index value a currency honestly and time it terribly. A currency stays mispriced longer than you can stay short. Valuation says where, momentum and positioning say when.
A cross like CAD/MXN is two USD legs in disguise, and relative volatility drives it, not macro. Weight the legs inversely to their vol so each contributes equal risk, not whatever the peso imposes.