FrançaisEvery day, dozens of banks publish price targets. Most are wrong, and following the average of opinions (« the consensus ») roughly tracks the index — while taking more risk for it. The real signal isn't « what do analysts say », it's « which ones were right, on what, and how much did it pay ».
We pull every timestamped analyst target (source: Yahoo Finance) and replay them against real prices: for each call we measure what it would have returned (long on a buy, short on a sell), winners AND losers. It's point-in-time: at each date we only use information available that day (no retroactive cheating). We then rank each analyst by real reliability — not « do they ride the rally », but « do they beat their sector » (alpha).
All start from the same pool: analysts reliable on the stock's sector (≥55% hit rate, beating their sector). Among them:
| Method | Return/yr | $100,000 grew to (2015→2025) | Volatility | Ratio |
|---|---|---|---|---|
| 🟢 JPI Light (prudent) | +21%/an | ~605 000 $ | ~25% | 0,84 |
| 🎯 JPI Fair (équilibré, recommandé) | +26%/an | ~946 000 $ | ~29% | 0,91 |
| ⚡ JPI Risk + (agressif) | +30%/an | ~1 350 000 $ | ~29% | 1,03 |
| Référence : S&P 500 | +12,6%/an | ~332 000 $ | — | 0,86 |
6 years (2015-2025), point-in-time, NVDA included, before fees and taxes. $100,000 starting. Volatility and ratio (risk-adjusted return) show the JPI Method is the best tradeoff.
📊 Explore the detailed backtests (interactive — 12 start months, 3 universes, stock by stock) →
The method leans toward beaten-down stocks (large gap to target). In the backtest, ~half the selected stocks had fallen over 12 months — and they rebounded +38%/yr on average the following year (vs +33% for those already rising). In other words: when a good analyst (reliable on the sector) is bullish on a stock everyone hates, that's often where the edge is.
This is not a sure thing. It's a single 11-year period, returns are irregular (driven by 2019/2023/2024, −23% trough in 2022 and −9% in 2018), before fees/taxes, and the universe only contains companies still in the index (limited but real survivorship bias). The JPI Method is a quality signal, not a guarantee. For the core of your portfolio, an index ETF stays rational; these formulas are a clear-eyed satellite. This is not personalized advice.
📊 Explore JPI Invest for free →
Rarely. Following the average of opinions roughly tracks the index, with more risk. The edge comes from isolating the rare reliable analysts by sector — which JPI Invest does by backtesting every call against real prices.
We replay every timestamped analyst target (2015-2025, point-in-time) against real prices and measure real profit per call. The JPI Method (median of ≥2 sector-reliable analysts) returned +26%/yr vs +12.6% for the S&P 500 — but over a single period, before fees, with no guarantee. See the detailed backtests, year by year →
The JPI Method (balanced) is the best risk/return tradeoff (ratio 0.91). JPI Light is more cautious (+21%/yr), JPI Risk + more aggressive and volatile (+30%/yr). The most reassuring signal: a stock present in all 3 lists at once.
JPI Invest aggregates analyst recommendations across the entire S&P 500 (plus the S&P MidCap 400), replays them against real prices and measures who predicts best — on results, not reputation. Instead of taking a price target at face value, you see each analyst's track record on each stock.