Why Analyst Price Targets Are Almost Always Wrong

This is because the model does not produce the target but the target produces the model.
Stock rallies 20%. Analyst raises target 20%. They call it revised assumptions. The conclusion came first and the spreadsheet got rebuilt around it.
If targets were honest, a price movement would only change the gap between where the stock trades and what it is worth. The intrinsic value itself would stay roughly constant until the business fundamentals actually changed. What you see instead is intrinsic value moving almost in lockstep with the price. That tells you where the analysis started.
Across the sell side, roughly 90% of ratings are buy or hold. Sell ratings average around 5%. Analyst careers depend on banking relationships, not forecast accuracy.
After 12 years investing my own money and six years working inside the industry, I noticed something counterintuitive. The targets that look most unrealistic are often the most mathematically honest. A $500 target on a $300 stock sounds delusional. But if the model was built forward from genuine assumptions rather than backward from a comfortable number, that is exactly what honest math sometimes produces.
The targets clustered suspiciously close to the current price are the ones worth questioning. Not the ones that look too bold!