When information becomes cheap, judgement becomes priceless

Like many professions, AI has transformed what's possible in investment analysis. Citadel founder Ken Griffin recently said AI has reached a "step change" in capability, allowing AI agents to perform in hours or days work that previously required teams of analysts with master's degrees and PhDs working for weeks or months.

Even I, in recent months, have used AI to synthesise large datasets, build stock-screening models and test investment ideas, sharing the results in various content pieces with the Livewire audience. 

This rapidly developing capability has changed the landscape and, in turn, created bigger questions. If every investor now has access to essentially the same tools, the same information and the ability to analyse it almost instantly, does AI make markets more efficient? And are there any unintended consequences?

According to Airlie Funds Management Senior Equities Analyst, Ray David, AI isn't just changing how investors research companies. It's changing how prices move, where opportunities emerge, and ultimately what skills will separate successful investors from everyone else. 

David believes the shift fundamentally changes where investment edge comes from. As information becomes increasingly commoditised, he argues the scarce resource is no longer data, but judgement.

"Judgement comes down to risk management... it's typically thinking about what's not priced in," David says. 

"You're trying to judge the risk about how everyone's positioned, what's currently factored in and what is not being factored in, what people are not thinking about."

In this interview, David explains why AI is accelerating the speed at which markets absorb information, why that is creating new distortions and opportunities, where active investors can still build an edge, and why judgement may become the most valuable skill in investing over the next decade.

The first casualty of AI may be one of investing's oldest edges

For decades, active investors have generated excess returns by processing information more efficiently and effectively than the market.

Sometimes that meant building better financial models. Sometimes it meant recognising a trend before everyone else. More often than not, it simply meant connecting dots that other investors hadn't yet joined together.

AI is changing that equation. The technology has dramatically reduced the cost of analysing public information. Company announcements can be summarised in seconds, earnings calls interrogated almost instantly and investment scenarios tested in a fraction of the time they once took. The productivity gains are obvious, but David believes the market is only beginning to grapple with the consequences.

One of the clearest examples is the way stocks now behave during reporting season.

Airlie's own research shows the average share price reaction to company results has increased markedly over recent years. 

Before 2020, companies typically moved around 3.5% on the day they reported. Over the past two years, that figure has almost doubled to around 6.5%.

Nobody can say AI is solely responsible. The rise of passive investing and systematic strategies has also changed market dynamics. But David believes AI is accelerating the speed at which information is reflected in prices, leaving investors with less time to exploit what has traditionally been one of the market's biggest inefficiencies.

He points to the so-called persistence effect, where companies that upgraded earnings often continued to outperform as analysts slowly revised their expectations higher. Likewise, disappointing results frequently led to a succession of downgrades as the market gradually accepted that conditions had changed.

For years, that lag created opportunity. David argues it is becoming increasingly difficult to exploit.

He cites research showing the persistence effect has almost halved since 2019, suggesting markets are pricing new information far more quickly than they once did.

"What's happening is that when an earnings downgrade comes out or an upgrade, the stocks are pricing in a much more savage reaction much more quickly," he says. 

"That alpha's being captured instantaneously."

Anyone who has watched recent reporting seasons will probably recognise the trend. Results that might once have produced modest moves now regularly trigger double-digit gains and losses. Whether AI is entirely responsible is open to debate, but it's difficult to argue that markets aren't reacting faster than they were five years ago.

For active managers, that raises an uncomfortable question. If everyone can process today's information almost immediately, where does tomorrow's edge come from?

The edge is becoming harder to find, but not impossible

David doesn't believe AI has eliminated the opportunity for active managers to outperform. If anything, he thinks it has made genuine differentiation more valuable. The catch is that investors need to look in different places.

"If everyone's using the same AI to read their reports," he says, "most first-hand information that's in the market is generally going to be priced in."

That pushes investors back towards activities that have always been central to good fundamental research but have arguably become less fashionable over the past decade -  speaking with customers, suppliers and competitors, travelling to visit businesses and developing proprietary datasets that can't simply be downloaded or scraped from public filings.

David points to Airlie's work on Pinnacle Investment Management (ASX: PNI), which reported recently, as an example. Rather than relying solely on consensus forecasts following Pinnacle's acquisition of Pacific Asset Management, one of Airlie's analysts travelled to the UK to meet management, advisers and competitors. 

Those conversations convinced the team the market was materially underestimating the earnings potential of the acquired business. The insight wasn't hidden in the annual report. It came from developing an understanding of the business that couldn't be replicated by reading publicly available information alone.

That may prove to be one of AI's more surprising consequences. As public information becomes easier to analyse, proprietary information becomes more valuable. David argues that doesn't mean active investing has become harder. It simply means the sources of competitive advantage are changing.

He believes successful investors will increasingly differentiate themselves in three ways: 

• by uncovering proprietary information the market doesn't yet have
• by applying deeper analytical judgement to the information everyone does have 
• and by maintaining a longer investment horizon than markets that are becoming increasingly focused on the short term.

The first of those is relatively intuitive. If AI is commoditising public information, original research naturally becomes more valuable.

Looking beyond the headline numbers

Finding information is only part of the challenge. David argues the second source of edge is analytical capability.

AI has become remarkably good at telling investors whether a company beat or missed earnings expectations. Where he believes experienced analysts can still differentiate themselves is understanding the quality of those earnings.

Was the result driven by stronger cash flow or aggressive accounting? Did profits improve because the core business performed better, or because an acquisition flattered the numbers? Is there something buried in the footnotes that materially changes the investment case?

Those questions require interpretation rather than information retrieval. As AI becomes increasingly effective at summarising what happened, David believes investors who can better explain why it happened, and what it means for future earnings, will continue to enjoy an advantage.

Judgement becomes the scarce resource

That idea runs through the entirety of my conversation with David. As information becomes increasingly commoditised, he believes judgement becomes the real source of competitive advantage.

That doesn't mean trying to predict the future. It means understanding what the market already believes, assigning sensible probabilities to alternative outcomes and recognising when the balance between risk and reward has shifted.

David illustrates the concept with a poker analogy. Holding a strong hand doesn't automatically justify pushing all your chips into the middle if there's a meaningful chance someone else is holding an even stronger one. Investing, he argues, demands the same discipline.

It's a subtle distinction, but an important one. AI can tell you what happened. It can even help explain why it happened. Judgement is deciding whether the market has already priced that information in, or whether something more important has been overlooked.

The final edge may simply be time

None of this should be mistaken for scepticism about AI itself. David is enthusiastic about the role it will play inside investment teams.

Airlie already uses AI to automate accounting analysis, interrogate company data and build internal tools that would previously have required months of manual work.

"We've built our own accounting quality integrity measurement system," he says. "This stuff is now being done in minutes rather than hours."

He also believes the technology will continue improving rapidly. Where he remains more cautious is around its ability to make investment decisions.

David recalls experimenting with AI by feeding it increasingly bullish and bearish research on the same company. Unsurprisingly, its recommendation shifted depending on the information it received.

For him, the exercise reinforced an important distinction.

"Ultimately the decision around pricing risk of a company comes down to the analyst," he says.

But perhaps the most overlooked source of edge has nothing to do with technology at all. 

It's time.

David argues that as markets become quicker to process information, they also become increasingly focused on the next result, the next earnings update and the next piece of news. Many quantitative strategies and multi-strategy hedge funds are forced to react quickly when positions move against them. Long-only managers with patient capital don't face those same constraints.

That creates an opportunity for investors prepared to think further ahead than the market. Rather than asking what a company will earn next quarter, David believes investors should spend more time thinking about what the business looks like 12 months from now, particularly when short-term volatility has pushed expectations too far in one direction.

Viewed through that lens, AI doesn't eliminate investment edge. It simply changes where investors need to look for it.

Public information is becoming increasingly commoditised, making proprietary information more valuable. As analysis becomes increasingly automated, interpreting what the numbers really mean becomes more important. And as markets become more reactive to short-term news, maintaining a longer investment horizon may become one of the few enduring advantages that can't be replicated by an algorithm.

When information becomes cheap, judgement becomes priceless.