From stock-tip newsletters to CNBC to YouTube, financial forecasting has always outrun its own accountability. A history of the problem and the tools that finally address it.
The confident stranger with a stock tip is not an invention of the internet. He has been a fixture of markets for as long as markets have had spectators, and every generation’s version of him has enjoyed the same convenient arrangement: his predictions are loud, public and profitable to make, while his track record is quiet, private and unprofitable to keep. The history of financial media is largely the history of this asymmetry – and of the very recent attempts to end it.
The newsletter era
Stock-tip newsletters flourished in America from the early twentieth century onward. For the price of a subscription, an editor with a typewriter and a tone of certainty would tell thousands of readers what to buy. The economics were beautiful for the editor: subscription revenue arrived regardless of outcomes, and a bad year could be buried by simply not mentioning it, since no reader had the time to audit years of back issues against price tables.
The pattern that would define every later era was already complete. Volume of predictions: high. Confidence: total. Systematic verification: none. When researchers finally did audit newsletter performance decades later – most famously in academic studies of the Hulbert Financial Digest data, which began independently tracking newsletter returns in 1980 – the results were what the incentives predicted: the majority underperformed simple market averages, and the loudest were no better than the rest. Hulbert’s project deserves its historical due as the first serious, sustained attempt at what this article is about: an outsider keeping the score the insiders wouldn’t.
The television era
Financial television, which matured in the 1990s, industrialised the pundit. Now the confident stranger appeared daily, in a suit, framed by tickers, answering “where do you see the S&P by year-end?” with a precision no honest person possesses. Television added two accelerants the newsletters lacked: real-time reach and the star system. A telegenic forecaster could become a celebrity, and celebrity, not accuracy, drove bookings.
The accountability tools of the era were occasional and artisanal. A columnist might revisit the year’s worst predictions each January; academics periodically published studies showing that televised buy recommendations, in aggregate, added little or nothing after their initial price pop. Philip Tetlock’s landmark research on expert political and economic judgment, published in 2005, gave the problem its scientific formulation: across thousands of tracked forecasts, prominent experts barely beat chance, and fame correlated with overconfidence rather than accuracy. The finding was devastating and changed almost nothing, because the audit was academic while the punditry was daily.
The YouTube era
Then the barrier to entry collapsed. By the late 2010s, anyone with a webcam could become a financial broadcaster, and the successful ones built audiences that rivalled cable segments – audiences bound to them not by a network’s credibility but by parasocial trust accumulated across hundreds of hours of intimate, direct-to-camera talk. The 2020-2021 retail trading boom poured rocket fuel on the format. Suddenly the confident stranger was in your pocket, spoke your language, showed you his own portfolio, and uploaded three times a week.
In one respect the new era was more honest than television: many creators disclosed their actual holdings and trades, something no anchor ever did. In every other respect the old asymmetry deepened. The volume of dated, specific, public financial claims exploded beyond anything the newsletter editors could have imagined – and the verification infrastructure remained exactly where it had been since 1910: nowhere. A creator could make three hundred stock statements a year, and the only memory of them was an unsearchable archive of video and the selective retrospectives the creator chose to film.
Why verification finally became possible
The strange thing about the YouTube era is that it created the conditions for its own audit. Unlike a 1955 newsletter or a 1998 TV hit, a YouTube call is permanently archived, precisely timestamped, publicly linkable and machine-transcribable. The claim, the date and the asset price are all public data. For the first time in the history of financial punditry, keeping the complete score is merely laborious rather than impossible.
Laborious still mattered – which is why the work waited for dedicated efforts rather than emerging spontaneously. The task is unglamorous: watch the videos, extract the exact quotes, timestamp them to the second, classify what kind of statement was made, pin the price on the date, track what followed, and refuse to delete the entries that age badly.
That work now exists. A new independent project, They Said Buy, applies it to the investing-YouTube ecosystem specifically: hundreds of dated quotes across dozens of creator-stock timelines, every quote linked to the exact second of the source video, with the subsequent price change shown against the S&P 500 over the identical dates. It is, in a direct line of descent, the Hulbert Digest of the creator era – an outside scorekeeper for a medium that never had one – with the crucial upgrade that readers can click through to the primary source and hear the words themselves.
The lesson of a century
A hundred years of financial punditry teaches one structural lesson: accuracy does not improve until someone else keeps the score. Newsletters did not audit themselves; television did not audit itself; creators will not audit themselves – not because the people are uniquely dishonest, but because no communications medium anywhere has ever volunteered for accountability that its economics punish. The corrections, when they came, always came from outside: a digest, a study, a tracker.
The difference today is that the outside correction can finally operate at the same scale and speed as the punditry it checks. The confident stranger is still here – he always will be, and honestly, markets would be duller without him. But for the first time since he started talking, the receipts are being kept, dated, linked and left in public view. What audiences do with that is the next chapter, and it is the first chapter of this history whose ending is not already predictable.