A former hedge fund manager runs 20 short-term strategies through real backtests and shows the actual numbers, win rates, and drawdowns instead of just describing them. The catch: most of the edges come from 1999 to 2003 Nasdaq volatility that no longer exists.
The one idea
Altucher's pitch is that most trading books describe ideas in vague terms (buy strength, sell weakness) without ever quantifying them. This book runs 20 specific, mechanical strategies through backtests on real data and shows the win rate, average profit per trade, drawdown, and Sharpe ratio for each one. The point is not any single system. It is the discipline of testing an idea before trusting it, and being willing to throw an idea out when the numbers do not hold up.
The strategies span gap fills, pairs trading QQQ against SPY, buying stocks under $5, a slowed-down version of the Turtle trend system, the 200-day moving average, option expiration effects, and a chapter that is explicitly about ideas that failed testing (gut feel, candlestick patterns, seasonality, P/E ratios). Almost every technique is built the same way: state a mechanical rule, run it on a specific universe over a specific date range, and report the result in a table.
The core method underneath all 20 chapters is more useful than any one of them: pick a testable rule, define the exact entry and exit, run it over years of data on a stated universe, and only trade it if the edge survives commissions and slippage. That process, not the QQQ gap-fill rule itself, is what a reader can still take from this book two decades later.
What it actually teaches
The book does not have one method, it has a repeatable testing process applied 20 times. Here is how Altucher builds and grades a single technique, using the gap-fill chapter as the template.
STATE THE RULE MECHANICALLY
No discretion. Example from Technique 1: buy a stock when it opens more than 2 percent lower than the prior close, sell at yesterday's closing price or at the close if that price is never hit. Every rule in the book is written as an if-then with a specific number attached, not a chart pattern you eyeball.
PICK A NAMED UNIVERSE AND DATE RANGE
Tests run on a stated basket, most often all Nasdaq 100 stocks including deletions (so failed companies are not survivorship-biased out), over a stated window such as January 1, 1999 to June 30, 2003. He reports the exact trade count, for example 9,821 trades for the basic gap-fill system.
REPORT WIN RATE AND AVERAGE RETURN, NOT JUST A CURVE
Every system gets a results table: percent winners, average profit percent on winners, average loss percent on losers, max consecutive losers, max drawdown. System #1 gap-fill: 62.87 percent winners, average trade 0.58 percent. He explicitly flags that 0.58 percent per trade is barely enough once commissions and slippage (he estimates up to 0.40 percent per trade) are subtracted.
STACK FILTERS UNTIL THE EDGE IS WORTH TRADING
Each system adds one condition and re-tests. System #2 adds a down day before the gap (0.75 percent average return, 4,938 trades). System #3 raises the gap threshold to 5 percent (1.97 percent average return, 993 trades). System #4 adds a market-wide gap filter (QQQ gapping down at least 0.5 percent), which on a $1,000,000 simulation from March 1999 to January 2003 produced 159.35 percent net profit, a 1.29 Sharpe ratio, and an 8.26 percent max drawdown over 525 trades.
TEST THE MIRROR CASE AND KEEP WHAT WORKS
He tries shorting gap-ups (System #5) and gets a negative edge, average return -0.56 percent across 752 trades, and reports that honestly instead of hiding it. He then tries holding the winning long setup overnight instead of closing same-day (System #6), which lifts the simulated return to 372.64 percent net profit with a Sharpe of 5 and a worse 11.89 percent max drawdown. The chapter ends by keeping System #6 and discarding the others.
What it looks like Monday morning
A reader who takes this book seriously does not copy the QQQ gap-fill rule into a broker's algo, that specific edge is a Nasdaq-bubble-unwind artifact and Altucher's own numbers show it thinning out even by 2003. What transfers is the workflow: write the entry and exit as an explicit rule, pick a real universe including the companies that later failed or got delisted, run it over a multi-year window, and report the win rate and average trade next to the drawdown before trusting it with money.
The other carryover is Technique 19, the chapter on what did not work. A reader should build their own version of that list: gut feel, chasing confirmation candles, candlestick patterns, calendar seasonality, and P/E-based long/short screens. Altucher tested all five and rejects them by name. That is a checklist of traps to test against before adopting any new idea, not just the ones in this book.
| System | Rule added | Trades | Win rate | Avg return/trade |
|---|---|---|---|---|
| #1 | Buy 2%+ gap down, sell at prior close | 9,821 | 62.87% | 0.58% |
| #2 | + prior day was a down day | 4,938 | 63.93% | 0.75% |
| #3 | 5% gap down instead of 2% | 993 | 60.93% | 1.97% |
| #4 | + QQQ also gapping down 0.5%+ | 525 | 61.14% | 2.07% |
| #6 | Same as #4, held overnight instead of same-day | 498 | 60.24% | 3.64% |
System #6's simulation on $1,000,000 from 1999-2003 shows 372.64% net profit but an 11.89% max drawdown, the tradeoff Altucher does not resolve for the reader.
But the 10th time you will lose everything. You. Will. Lose. Everything.James Altucher, Trade Like a Hedge Fund
Where it fails
- Almost every backtest window ends in 2003, inside the exact regime that made it work. The flagship gap-fill and swing-gap systems are tested from March 1999 (QQQ's inception) to January 2003, which captures the dot-com blowoff and the 2000-2002 bear market crash, the highest-volatility, most gap-prone stretch in modern Nasdaq history. Altucher's own equity curve chart shows the strategy spiking hardest during the market's worst months. Nasdaq 100 daily gap frequency and size fell sharply after 2003 as the stock moved from a speculative-mania name-set to a mega-cap-dominated index; there is no evidence in the book, because it could not exist yet, that these edges survived the calmer, more efficient markets of the following two decades.
- The book assumes access to a professional backtesting platform and historical tick data most readers do not have. Every result table cites Wealth-Lab, Inc. software and a maintained database of Nasdaq 100 constituents including deletions (delisted names, essential to avoid survivorship bias). A retail reader without that data feed and platform cannot reproduce a single test in the book, they can only take Altucher's tables on faith, which cuts directly against the book's own stated point that you should never trust an edge you have not verified yourself.
- Several 'successful' systems are barely profitable after costs, and the book says so, but still presents them as techniques. System #1 (basic gap fill) returns 0.58 percent per trade against an admitted 0.40 percent commission and slippage drag, and System #2 nets 0.75 percent against the same drag. These are included as numbered techniques in the table of contents alongside the strategies that cleared that bar by a wide margin, so a reader skimming for method names rather than reading the actual result tables can walk away with a false sense of how many of the 20 techniques are worth trading.
- The 200-day moving average chapter undercuts the crossover signal financial media still repeats today. Altucher's own test of the standard 'buy when price crosses above the 200-day moving average' rule shows only 40 of 142 occurrences profitable since 1950 (28 percent), and since January 2000 only 2 of 17 occurrences worked. He still recommends a variant (buy when price is 20 percent below the 200-day average, hold one month), but that system requires the reader to already have decades of index data and the willingness to buy into a falling market, which is a much harder trade psychologically than the book's one paragraph on it suggests.
- The Turtle-style trend system is credited to a manager who told Altucher about it verbally, with no published track record shown. Technique 7 states the 22-week over 55-week moving average crossover rule came from 'a manager of a multibillion-dollar trend-following fund' but does not name the fund, show its audited returns, or provide the same trade-by-trade table the other 19 techniques get. It is the one technique in the book presented on authority rather than on the tested-evidence standard the rest of the book insists on.
- Several techniques depend on market structures that later changed or vanished. The QQQ-SPY pairs trade (Technique 2) and the deletions-from-the-indexes trade (Technique 10) both depend on specific index-rebalancing mechanics and correlation relationships from the early 2000s. Index fund assets under management, rebalancing rules, and the number of funds arbitraging index adds and deletions all grew substantially after this book was written, which is exactly the kind of crowding-out effect Altucher himself warns about in Technique 19 when he describes seasonal edges disappearing once enough people trade them.
Who it is for
Buy it if
You want a worked example of how to actually backtest a short-term trading idea, win rate, average return, drawdown, sample size, rather than another book that describes chart patterns in prose. Useful as a template for building your own testing discipline, especially the Technique 19 list of things Altucher tested and rejected.
Skip it if
You are looking for a strategy to trade today. Every specific numeric edge in the book is measured on 1999-2003 Nasdaq data and none of it is re-verified against current markets. Skip it if you do not have backtesting software and historical data, since you cannot check any of the claims yourself, which is the exact failure mode the book warns against.
As an Amazon Associate I earn from qualifying purchases. It costs you nothing extra, and it does not change what gets recommended or what the verdict says.