The method: spotting analysts with a real edge
The goal, the backtest, and the formulas that work — with the numbers.
The JPI method
On average 2× the S&P 500 return since 2015.
1
We rate analyst performance
2
Predictions compared each year since 2015 to real results
3
We keep the best per stock and sector, the rest excluded
The JPI method in one sentence: we analyze the analysts (not the stocks) → we replay all their targets against real prices → we keep only those who beat their sector → we backtest it over 11 years → and we show you, stock by stock, who was actually right — and therefore who is credible for their next calls and estimates.
In short — Most analyst targets are wrong, and « the consensus » almost never beats the market. JPI Invest replays every analyst call against real prices, sector by sector, and keeps only the rare ones with a real edge. On a 2015-2025 backtest (point-in-time, NVDA included), following the JPI Method returned +19.8%/yr vs +11.5% for the S&P 500.
The result: a bullshit detector. Instead of following « the consensus » (which almost never beats the market), you see the rare analysts with a real edge, by sector, with backtested numbers.
The problem
Every 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 ».
How we measure
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).
The 2 formulas that work
All start from the same pool: analysts reliable on the stock's sector (≥55% hit rate, beating their sector). Among them:
- 🎯 JPI Method (recommended) — require ≥2 reliable analysts bullish and take the median of their targets. Balanced and robust.
- ⚡ JPI Risk + — among the reliable ones, follow the most optimistic. Aggressive, more volatile.
| Method | Return/yr | $100,000 at the start, all sold and put back in each year (2015→2026) | Volatility | Ratio |
|---|---|---|---|---|
| ⚡ JPI Risk + | +22,1% | 899 010 $ | 24% | 0,93 |
| 🎯 JPI Method | +19,8% | 728 916 $ | 21% | 0,93 |
| S&P 500 | +11,5% | 332 028 $ | 15% | 0,79 |
11 years (2015-2025), point-in-time, NVDA included, before fees and taxes. Start with $100,000; every year everything is sold and the full amount — capital plus gains, or minus losses — goes into the new selection. No fresh money is ever added, and nothing is left untouched for 11 years either. Ratio = risk-adjusted return (return/volatility): JPI Risk + is the highest (1.02), while JPI Method (the default recommendation) favors robustness — median of at least 2 analysts.
📊 Explore the detailed backtests (interactive — 12 start months, 3 universes, stock by stock) →
A telling example
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.
Being honest
This is not a sure thing. It's a single 11-year period, returns are irregular (driven by 2019/2023/2025, down years in 2018 and 2022), before fees/taxes, and companies that left the indices have no analyst history on our side, so the method could never pick them (a residual bias we can't measure). 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.
🔧 Under the hood (what we can show)
Two clocks run in parallel, and that's the heart of the method: ① to judge an analyst, we look at their entire history of matured calls (10+ years of hindsight today); ② for today's selection, we only take their targets from the last T days — what they think now, not two years ago.
A firm is deemed reliable on a sector if:
- it has at least N matured targets in that sector (enough sample)
- its win rate is above K % (it beats a coin flip)
- it is active every year (not a ghost firm)
We then keep only its most recent view, with a credible target and upside of at least M %.
From the retained stocks, the 2 formulas: 🎯 Fair = the median target of at least 2 reliable firms (recommended) · ⚡ Risk+ = the most optimistic (the highest bet). A guardrail completes it: Fair requires a stock with enough track record — fresh IPOs (e.g. SpaceX) only make Risk+.
🔒 The exact thresholds are our secret sauce. But nothing is hidden on substance: it's all backtested over 11 years and verifiable on results, not on trust.
📊 Explore JPI Invest for free →
Frequently asked questions
Does analyst consensus beat the market?
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.
What is the JPI Invest backtest?
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 +19.8%/yr vs +11.5% for the S&P 500 — but over a single period, before fees, with no guarantee.
Which formula is best?
It depends on your profile. On the backtest, JPI Risk + has the best return (+22.1%/yr, beating the index 10 years out of 11) — but it's aggressive and volatile. JPI Method (the default recommendation) is the most robust: median of at least 2 reliable analysts (+19.8%/yr). The most reassuring signal: a stock present in both lists at once.
What is JPI Invest?
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.
- 🏆 Analyst leaderboard by real reliability (beating their sector, not just riding the rally)
- 🎯 What to buy: 3 backtested methods over 11 years, from cautious to aggressive
- 💬 AI assistant: ask anything about any stock, with the numbers
- 📊 Portfolio tracking + 🔔 price alerts by email