For 30 years, my advantage as an investor was painstaking research. AI just made it worthless.
In 1995, I read two books that changed my life: Benjamin Graham’s The Intelligent Investor and Roger Lowenstein’s biography of Warren Buffett. I was hooked on value investing. I could not get enough of it. In those days, however, getting enough of it was hard. It was the infancy of the internet, so it wasn’t as if I could just Google “Warren Buffett” or “value investing.”
The first thing I could get my hands on was Berkshire Hathaway’s annual report. I had to look up the phone number in the phone directory. After that, I placed a phone call using a landline. A nice lady with a Midwestern accent picked up the phone, “Berkshire Hathaway — how can I help you?”
“Hello,” I said. “I’d like to order a copy of your annual report.” That’ll be no problem, came the answer. I gave her my mailing address, and she repeated it back to me. A few days later, I received a brown envelope with the annual report. I devoured it. And I used the table of Berkshire’s publicly traded investments to go after new materials — the reports of Coca-Cola, Geico, American Express and Capital Cities/ABC — in a similar fashion, by picking up a landline phone and making a verbal request.
What’s my point?
Things went slowly. Acquiring knowledge was hard. Every step in the acquisition of investing knowledge involved a cycle of days, if not weeks. The various pieces had to be carefully assembled and assiduously harvested.
From those early days, I learned about the Berkshire Hathaway meetings. They were not recorded for YouTube or televised live. So, I acquired some shares and attended them in person. All in pursuit of an investing edge. Yes — there were a few lucky non-attendees of those meetings who managed to get onto Whitney Tilson’s famous email list — which at the time numbered dozens at most. But that was it. The inside tracks were few and far between.
Because I could not get enough, I also attended the Wesco meetings in Pasadena, California with Charlie Munger, and the Sequoia Fund meetings run by Bill Ruane and Rick Cunniff. Once at the Wesco meeting, Munger mentioned the Tupperware company in the context of Robert Cialdini’s Influence:The Psychology of Persuasion. I was on it. I even organized a Tupperware party in my home to understand how the company’s sales system worked.
In those days, potential investors in my fund used to ask, “What’s your edge?” Perhaps it was a shorthand for “Do you have a source of inside information?” Or, more politely, “What’s your variant insight?” If I had any at all, it was based on this sort of scuttlebutt.
But that world is now gone.
Over the last decade many of those advantages have been eroded. Pounding the pavement, smiling and dialing were no longer necessary. Because those hard-earned insights were being emailed, tweeted, livestreamed, YouTubed, podcasted and more. Yes, you still had to pick your sources. But there was so much more that was simply available at the tips of any analyst’s fingers.
And now comes AI. The way investment research changed in the internet age was glacially slow when compared to the earthquake that is LLMs. Before ChatGPT, you still had to have the patience to assemble and read through the mosaic of sources. Now you can just ask ChatGPT or Gemini to do all the trawling. An LLM can give an instant summary of all that has been said in the public domain about a company. And it can be instantly analyzed to provide the current state of wisdom on the topic.
Granted, the LLMs cannot, to the best of my knowledge, generate original thinking. But the path to original thoughts is so much shorter — and so much more accessible.
In past times, I used to make a point of traveling to the Kings Road in London on the off chance that Nick Sleep and Qais Zakaria of Nomad might treat me to a Cornish pasty along with some of their original thoughts. It was there that I first heard Nick talk about “scale economies shared.” Today there is no need for me to travel all that distance — unless I am only after the Cornish pasty and Nick and Zak’s fine company.
Why? So many original thoughts or variant opinions are available online and the LLMs can trawl through all of them — adding in and cross-referencing a mosaic of additional sources: social media posts, SEC filings and more. That material is all now public domain.
It’s true that there are still many nonpublic data sources that are not accessible by LLMs. But data tends to leak out via Scribd, Reddit, arXiv or social media. Even if that’s just metadata, it can be assembled effortlessly. Then a human like me can apply intuition and fact-check the sources to make sure the LLM is not hallucinating.
What does this mean for value investing?
My conclusion is that the golden age of value investing is well and truly over. Thanks to the LLMs, I have no need for a junior analyst: At a fraction of the cost, an LLM can do a far better job. And the era of the swashbuckling hedge fund manager — Michael Steinhardt, George Soros, Julian Robertson — is also gone.
Asset pricing will become more accurate. Hidden subtleties about a particular business that make it either better or worse than it appears will be more easily revealed. The return on better analysis and better insight will diminish because it will be available to all. The British scientist and writer Matt Ridley would liken this to the Red Queen effect: We can all run faster, but at the end of the day, we will all have access to the same LLMs. The dispersion of returns for active management has to reduce, clustering even closer to the index return, even for investors who don’t closet index.
People used to ask, “What’s your edge?”, but there is no edge anymore. At least, none that is obvious to me.
Where, then, will money flow? What remains are index funds and asset-gatherers like BlackRock. And what of small, independent boutique funds like mine? We can stick to human behaviors that work — which the indexers and the quant funds with their large organizations and programmed behavior cannot replicate. We can buy and hold, which is sometimes called time arbitrage.
I will certainly continue to invest in scuttlebutt, LLM channel checks, expert calls and the like. But I know that it will probably be futile. Because everyone can do it. I expect my attention to turn to building the best set of relationships, and to invest in businesses that themselves invest in better sets of relationships.
For this I can think of no better example than Berkshire Hathaway — with its vast number of shareholders. That’s a large group who have the lived experience that buying and holding for the long term, rather than chopping and changing, is the way to invest. That will give Buffett’s successor, Greg Abel, a far better environment to allocate capital — just as my investor base should afford me the same. I know it’s slim pickings, but it should be enough for those of us who remain.
If the only thing AI and LLM's do for finance is to level the playing field — to further marginalize the masters of the universe (and wannabees like me) — would that be such a bad thing? Most likely not.
Thanks to AI, the playing field is more level than it's ever been. Now more than ever there is no need for the average investor to have to pay outrageous fees. A big benefit of the stock market was democratize finance. Perhaps, now, thanks to AI, that day is finally here.
A message from Advisor Perspectives and VettaFi: Discover something new! Click here to register for our upcoming webcasts.
Bloomberg News provided this article. For more articles like this please visit
bloomberg.com.
Read more articles by Guy Spier
