A Field Guide
Misreporting in
investment services.
The Tide Picker founder has investigated dozens of investment services to see if their performance claims could be verified. In the vast majority of cases they could not — undone by sleights ranging from the subtle to the outrageously brazen.
What follows is a field guide. None of these tricks are theoretical: every one has shown up, repeatedly, in services that charge real money for their signals.
Impossible-to-earn pricing
Cherry-picking the day's low as the buy price and the day's high as the sell price. No real trader hits both ends of the daily range, and assuming you do is wishful thinking that will be wrong 100% of the time in reality. Honest backtests use closing prices or, at minimum, conservative fills like the volume-weighted average price.
Dropping inconvenient trades
Some services quietly ignore losing positions, reporting only the winners. Others retroactively reclassify a bad trade as a long-term hold or a different strategy entirely. A clean backtest shows every signal the system would have generated, including the ones that lost money.
Summing returns instead of averaging them
Buying 10 stocks that each gained 100% is a 100% return on the basket, not 1,000%. Some services add the percentages together to inflate the headline. Real returns are risk-weighted: dollars in, dollars out, averaged across the capital actually deployed.
Look-ahead bias
Using information in the backtest that wouldn't have been available when the trade was placed — restated earnings, revised economic data, future index reconstitutions. The model effectively trades with tomorrow's newspaper.
Cherry-picked start and end dates
Beginning a track record at the bottom of a bear market and ending at a bull-market peak makes almost any strategy look brilliant. Shifting the start date by a few months often erases most of the alpha.
Hiding drawdowns
Reporting only annualized returns and never showing the worst peak-to-trough loss. A 30% CAGR that requires sitting through a 70% drawdown is a very different product than one that drew down 20%.
Case studies
See these tricks in the wild
Specific postmortems on real services and strategies — including the math that proves the claims couldn't be real.
How Tide Picker reports
Every backtest on this site uses closing prices for both entries and exits, includes every signal the system would have generated — winners and losers alike — and is measured across the full backtest window without cherry-picked endpoints.
Returns are reported as compound annual growth alongside maximum drawdown, volatility, Sharpe, Sortino and Calmar ratios, with trading costs and financing charged against the results. Where a figure is an estimate rather than a measured result — such as the blended return of holding both books in equal weight — we say so on the page. No backtest can predict the future, and we may simply be wrong in the long run. But the numbers are what the historical data showed, reported honestly.