1. A stock drifts up 2 percent across a quiet week with no filings or announcements, then jumps 6 percent the day an 8-K discloses a buyout offer. Which move can a systematic process be built around, and why? A The drift, because slow moves are easier to catch B The 8-K move, because it is larger C The 8-K move, because it belongs to a recurring, dated, disclosed class of events that spans the whole market and can therefore be studied as a population D Both equally, since both moved the price
2. What separates an event-driven signal from a chart pattern? A Event signals update in real time B Events are scheduled and force constrained participants to transact, creating structure that isn't just noise in past prices C Chart patterns only work on intraday timeframes D Event signals don't require historical data
3. A stock is added to the S&P 500. Who is the structurally constrained participant? A The company's insiders B Index-tracking funds, which must buy on the effective date regardless of price C The market makers quoting the stock day to day D Retail traders reacting to the headline
4. A backtest shows that stocks whose tickers start with a vowel outperformed for five years. A colleague argues the sample is large and the effect is statistically significant. What is the mechanism-first objection? A Five years is always too short a sample B Statistical significance can only be established with ten or more years of data C Nobody is forced or even incentivized to trade based on ticker spelling, so there is no reason the pattern should exist, no way to know when it dies, and every reason to suspect noise D The strategy would be too expensive to trade
5. An event class shows a reliable 4 percent downward drift over 60 days, but the affected names cost 25 percent annualized to borrow. What does mechanism-first honesty say about shorting it? A Take the trade: a reliable drift is a reliable edge B The edge is roughly a wash: 60 days of borrow at 25 percent annualized consumes about 4 percent, so the carry eats the drift C Borrow costs don't matter because they are paid at the end D The drift must be fake if borrow is that expensive
6. Your study of a queryable event returns a positive average drift from 35 events, all occurring in one unusually volatile year. Following the four-step loop, what is the correct next move? A Trade it small while collecting more data B Restate the hypothesis so it only claims the effect for volatile years C Treat the result as unproven: the attack step flags small sample size and single-regime concentration, so the study needs more history and a regime split before it is evidence D Publish the result, since the mean is positive
Check answers