Every ballplayer knows the sinking feeling of watching a routine grounder skip through the infield. That is exactly what software investors have experienced over the past year, watching names like Adobe, Salesforce, and Workday get chopped in half while the S&P kept rounding the bases without them.

The story fits on a bar napkin. AI coding agents can supposedly spit out software for next to nothing, so why would anyone pay premium multiples for companies whose products appear to be little more than code? Sell first. Ask questions later.

Sparkline Capital recently published a paper examining this very question, and it is worth your time even if you have never owned a software stock. The framework beneath it extends far beyond technology. It applies to every value investor who has ever bought something that looked cheap only to watch it get even cheaper.

Cheap Is Not the Same as Safe

Sparkline’s research team revisited some of the ghosts of value investing past: Blockbuster, Borders, RadioShack, and McClatchy. Every one of them looked like a bargain right up until the day it wasn’t. Revenue often lagged the collapse in the share price by several years, creating the illusion of cheapness. Value investors piled into seemingly attractive valuations and rode those ships straight to the bottom.

That is the danger lurking beneath today’s software selloff.

GoDaddy, Adobe, Workday, Atlassian, and Salesforce now trade at roughly half the market’s forward earnings multiple. On a traditional Graham-style scorecard, that looks like Christmas morning. Yet Sparkline’s backtests, stretching all the way back to 1994, show that plain-vanilla value metrics—price-to-earnings, price-to-book, and the usual suspects—have consistently struggled in industries facing genuine technological disruption. Since 2010, those metrics have actually produced negative results in disrupted industries even while continuing to work well everywhere else.

Cheap stocks in disrupted industries are not automatically bargains. More often than not, they are simply businesses whose balance sheets have not yet caught up with reality.

The Moat Was Never the Code

Here is the part that matters for stock picking.

Sparkline developed what it calls an Intangible Value factor—essentially Ben Graham’s toolkit updated for the modern economy. Instead of focusing solely on tangible book value, the model also gives credit for brand equity, human capital, network effects, customer relationships, and intellectual property.

Run that framework across today’s software wreckage and the picture changes dramatically. Many of these beaten-down companies actually screen as attractive once you account for the economic moat sitting beneath the ticker symbol.

Why?

Because for most of these businesses, the code was never the moat.

Adobe does not survive simply because it writes elegant software. It survives because generations of designers learned Photoshop first, because enterprise customers trust the brand, and because moving an entire creative department off Creative Cloud is a multi-year undertaking that few organizations want to own.

Salesforce is not merely a CRM platform. It is a system of record embedded throughout thousands of customer workflows, supported by an enormous ecosystem of consultants, developers, and implementation partners. Replacing that infrastructure is far more difficult than a polished demo video might suggest.

This echoes economist David Teece’s classic theory of complementary assets.

EMI invented the CT scanner but ultimately lost the market to General Electric because GE already possessed the sales force, service network, financing capability, and hospital relationships.

RC Cola introduced diet cola before Coca-Cola, yet Coke still won because it controlled the shelf space and distribution.

The lesson repeats itself in every technological cycle.

The innovator does not always win.

The company that owns the distribution, customer relationships, switching costs, trusted brand, and complementary assets usually does.

Where the Real Doubles Hide

Sparkline’s most compelling insight lies at the intersection of two characteristics: companies aggressively embracing artificial intelligence while simultaneously possessing durable competitive advantages that extend well beyond AI itself.

Names like Salesforce and Adobe land squarely in that upper-right quadrant.

By contrast, companies such as Duolingo appear to fit the definition of a classic value trap—stocks that look inexpensive on the surface but lack the deeper competitive advantages needed to justify long-term optimism.

The broader insight, however, involves dispersion.

This is the part that should get every value investor’s pulse racing.

Sparkline created a basket it calls Disruption Scare Stocks—companies that have declined at least 30% while operating in industries facing meaningful technological disruption.

The median 12-month return is fairly ordinary, roughly matching the broader market.

The tails tell a completely different story.

Ten percent of these stocks double over the following year.

Sixteen percent lose half their value.

The spread between winners and losers is nearly twice that of the average stock.

That is the entire ballgame.

When outcomes become that widely dispersed, even a modest edge in separating future winners from future losers becomes extraordinarily valuable. You do not need to be right every time. You simply need a repeatable framework that helps distinguish tomorrow’s New York Times from tomorrow’s Blockbuster.

The reward for making that distinction correctly has rarely been greater.

Why Insider Buying Belongs in the Toolkit

Sparkline’s research ends with intangible value.

I would add one more tool to the framework—one that has been producing useful signals for decades.

Insider buying.

Think about what an intangible moat really represents.

It is trust.

It is customer relationships.

It is institutional knowledge.

It is switching costs.

It is everything that fails to appear on a traditional balance sheet.

So who has the best information about whether those competitive advantages remain intact?

Not the analyst updating a discounted cash flow model from his kitchen table.

The people running the company.

The CFO watching renewal rates.

The founder who knows whether enterprise customers are expanding contracts or quietly looking for the exits.

The academic evidence is not folklore.

Lakonishok and Lee found that insider purchases carry a meaningfully stronger predictive signal than insider sales. Jeng, Metrick, and Zeckhauser estimated that insider purchases have historically outperformed the broader market by several percentage points annually. Nejat Seyhun spent an entire career documenting essentially the same pattern across decades of market data.

None of this is secret information.

It is simply underused because most investors treat a Form 4 filing the same way they treat a box score from a game whose outcome they already know.

Insider selling tells you very little.

Executives sell for dozens of perfectly ordinary reasons—a child’s tuition, estate planning, diversification, taxes, or a trading plan established months earlier.

Insider buying is an entirely different animal.

Cluster buying—multiple executives or directors committing meaningful amounts of their own capital after a stock has already been cut in half—is one of the strongest signals investors can find.

Nobody willingly catches a falling knife with their own paycheck unless they believe the business underneath is considerably stronger than the market assumes.

Now overlay that signal on Sparkline’s framework.

A Disruption Scare Stock with a strong Intangible Value score is already interesting.

The same company with insiders quietly buying into the drawdown becomes something even better—a second, completely independent vote of confidence coming from the people with the best seat in the house.

When both signals point in the same direction, you are looking at something very close to a value investor’s version of a called third strike.

The pitch was right there.

The software selloff is neither a blanket bargain nor a blanket trap. It is a dispersion machine, producing both future Blockbusters and future New York Timeses from the same pool of companies, while the market does a poor job distinguishing between them.

Traditional value metrics alone are not enough.

Intangible value—the kind that recognizes the moat surrounding a business rather than simply the furniture inside it—has a demonstrated track record of separating likely winners from likely losers.

Layer insider buying on top of that and you gain a second opinion from the people who know the business best.

That combination—a company trading cheaply on an intangible basis while insiders are committing meaningful personal capital—is about as close as investing gets to seeing a fastball down the middle.

Swing at those.

Take a pass on everything else.

Here are 4 software stocks with recent insider buying that may be worth considering:

Adobe (ADBE)

Adobe possesses perhaps the widest moat in today’s software landscape, and the insider tape has begun to acknowledge it.

Director David Ricks purchased 10,000 shares in the open market on June 25, 2026, at a weighted average price of $194.51, increasing his personal holdings by roughly 96%.

That said, investors should not mistake one director’s conviction for broad insider enthusiasm. Insider selling has exceeded insider buying over the past three months, including a significant sale by the CEO. Think of this as one meaningful vote of confidence rather than broad insider conviction.

Fiserv (FISV)

Fiserv is the one name on the list that genuinely resembles a textbook cluster-buy.

Over a single week in mid-June, three different insiders reached for their own checkbooks.

CFO Paul Todd purchased 10,060 shares for approximately $500,000.

Chief Administrative and Legal Officer Adam Rosman bought 10,150 shares at $49.33 per share.

Director Wafaa Mamilli added another 2,960 shares.

In total, eight insider purchases have appeared over the past year without a single insider sale—a remarkably clean signal, particularly for a company trading at roughly eight times earnings.

That combination deserves attention.

Autodesk (ADSK)

Autodesk offers one of the strongest insider buying patterns of the group.

The CEO, CFO, and two directors all purchased shares within a matter of weeks.

CEO Andrew Anagnost bought 2,460 shares at $202.66 per share.

CFO Janesh Moorjani purchased 2,500 shares just two days earlier.

Directors Stacy Smith and John Cahill each followed with purchases worth several hundred thousand dollars.

When both the CEO and CFO are signing Form 4s within the same month, that is not tax planning.

That is management telling you they believe the market has materially undervalued the business.

Skillsoft (SKIL)

Skillsoft is easily the speculative name on this list.

The company has endured a difficult stretch, and the insider buying resembles a calculated bet on a turnaround more than a vote of confidence in an already healthy business.

Director Jim Frankola purchased 50,000 shares valued at roughly $302,730, more than doubling his position. The stock responded by jumping 17% the day the filing became public.

Investors should recognize the risks.

Skillsoft continues to report widening losses, declining revenue, and share-price volatility greater than roughly 90% of publicly traded U.S. companies.

Treat this as a speculative turnaround supported by one director’s conviction—not the kind of broad, institutional-grade insider buying currently visible at companies like Fiserv or Autodesk.