Technology markets have a way of rewarding investors who make big, concentrated bets. After a decade in which a relatively small group of technology companies produced outsized returns, concentration can start to look like conviction while diversification looks like a concession. According to Neil Druker, Managing Member of Melanie Lane Capital Management in Boston, the two ideas aren't interchangeable. Owning exceptional companies and building a portfolio that can withstand different market conditions requires distinct yet related decisions.
That distinction becomes most obvious when the market changes.
A manager can be right about a company's product, competitive position, and management team and still end up with a portfolio that's vulnerable to a slowdown in enterprise spending, tighter financing conditions, or a shift in investor expectations. Good research doesn't automatically protect a portfolio from those risks. Portfolio construction is where those risks have to be considered before they become obvious.
Concentration and Conviction Are Not the Same Thing
Concentration can happen for good reasons. A manager may identify a handful of unusually attractive opportunities and decide they deserve larger positions. But concentration can also develop without anyone realizing it because several holdings depend on the same underlying assumptions.
That's particularly easy to miss in technology.
Companies that appear unrelated based on their products may all depend on enterprise software budgets, cloud spending, digital advertising, semiconductor demand, or inexpensive financing. A portfolio can be spread across 10 different businesses and still amount to a single large bet on the same economic condition.
That makes identifying hidden correlations across technology holdings an important part of portfolio construction. The question isn't simply how many companies a portfolio owns. It's how many different things those companies actually depend on.
Company Quality Is Not Investment Quality
A strong company isn't automatically a strong investment at every price.
A business can have excellent management, a large addressable market, strong unit economics, and a competitive advantage that looks durable. If investors have already priced in years of exceptional execution, however, the stock may still produce an ordinary return.
That's the distinction between company quality and investment quality.
Separating business quality from investment quality helps investors keep those two judgments apart. It's possible to admire a company, believe it will remain successful, and still decide that its stock isn't attractive enough at the current valuation.
That discipline becomes particularly important in technology, where enthusiasm around a company's future can sometimes move faster than the underlying economics.
Valuation as a Framework for Assumptions
Valuation can become difficult in fast-moving technology markets because the assumptions behind a forecast can change quickly. The problem isn't valuation itself. It's the false precision that can come from relying too heavily on a single forecast or price target.
Scenario analysis provides a more useful way to think about the question.
Instead of asking exactly what a company is worth, an investor can ask what would need to happen for the current price to make sense. How quickly would revenue need to grow? What margins would the business eventually have? How much capital would it need to support that growth? How long would its competitive advantage need to last?
Writing those assumptions down makes them easier to challenge.
It also helps investors see how much room the current valuation leaves for something to go wrong. A stock doesn't necessarily need to be dramatically overvalued to produce a disappointing return. Sometimes the problem is that the price assumes too much goes right.
Position Sizing Is Where Conviction Meets Fiduciary Humility
Position sizing shouldn't be based solely on conviction. It also has to reflect what happens if that conviction turns out to be wrong.
That means looking at liquidity, volatility, balance-sheet risk, potential dilution, and how closely the position moves with other holdings. A company may warrant a place in the portfolio while still meriting a moderate position if the downside is difficult to estimate or the valuation depends on a narrow set of assumptions.
The reverse can also be true.
A less exciting business with a broader range of reasonable outcomes may justify a larger position even if its upside case doesn't generate the same enthusiasm.
For Neil Druker, that tension is part of the discipline of institutional investing. Conviction matters, but so does the responsibility that comes with putting other people's capital behind an investment thesis.
Diversification That Protects Judgment
Diversification is sometimes criticized for diluting the benefit of good research. Poorly designed diversification can certainly do that. Owning dozens of securities simply to make a portfolio look prudent doesn't necessarily reduce meaningful risk.
But the argument against diversification often assumes something no investor can reasonably claim: that research eliminates uncertainty.
It doesn't.
Technology markets change quickly. Management teams make mistakes. Competitive advantages weaken. New products can emerge from places that weren't obvious when an investment was made. Even a well-researched thesis can turn out to be wrong.
Thoughtful diversification creates room for those mistakes without allowing one forecast, management team, or market narrative to determine the outcome of the entire portfolio.
The goal isn't to eliminate conviction. It's to make sure conviction doesn't become fragility.
Liquidity Is an Investment Variable
Liquidity is often treated as an operational issue, but it can directly affect investment outcomes.
A manager’s ability to adjust a position when circumstances change can be an important consideration in portfolio management. That becomes particularly important when markets are moving quickly, and several holdings could be difficult to sell simultaneously.
Before building a position, investors should consider how long it would take to reduce it under normal and stress conditions. They should also look at whether several holdings share the same potential liquidity problem.
The same question applies at the fund level. If investors expect to have access to their capital on a relatively short timetable while the underlying assets may take much longer to sell, the mismatch can create pressure at exactly the wrong moment.
A temporary decline can become a permanent loss when a manager has to sell because of a liquidity constraint rather than because the investment thesis has changed.
Looking Past the AI Label
Artificial intelligence has made portfolio construction even more complicated because the label itself doesn't tell investors very much.
One company may benefit directly from demand for computing and data-center infrastructure. Another may benefit from increased productivity. A third may be selling an AI product amid intense competition and rising costs, making its long-term economics less attractive than the headline suggests.
The more useful question is what the portfolio actually depends on.
How much expected value depends on continued AI capital spending? How much depends on rapid enterprise adoption? Which companies need favorable regulation? Which ones need to demonstrate that AI usage can translate into sustainable margins?
Those dependencies can overlap across holdings even when the companies themselves look unrelated.
A portfolio may appear diversified while several positions are ultimately tied to the same AI spending cycle.
The Mistake Strong Markets Encourage
One of the easiest mistakes to make during a strong market is assuming that because risk hasn't materialized, it must have declined.
During sustained rallies, correlations can look favorable, liquidity feels plentiful, and valuation discipline can start to seem overly cautious. When everything is working, investors have less incentive to ask what happens if the environment changes.
That's exactly when those questions matter most.
Testing portfolios against unfavorable assumptions is much easier when the market is calm than when a repricing is already underway. A portfolio can't be made resilient in the middle of a shock with the same degree of control that was available beforehand.
Resilience as the Actual Objective
The goal of portfolio construction isn't to eliminate volatility. Investors accept volatility because participating in markets can create meaningful long-term returns.
The more practical goal is to build a portfolio that any single mistake, narrative, or market environment cannot permanently damage.
For institutional investors in technology, that means looking beyond the question of which companies have the highest expected returns. It also means asking which assumptions the portfolio relies on, where those assumptions overlap, and what happens if one of them proves false.
Neil Druker's framework isn't a prescription for how much technology exposure an institution should have. It's a way of thinking about that exposure more deliberately: understand the businesses, understand the assumptions behind their valuations, size positions with both opportunity and downside in mind, and build enough resilience to remain disciplined when the market stops cooperating.
This article is educational and analytical in nature. It does not constitute investment, legal, or tax advice, does not recommend any security or transaction, and is not an offer or solicitation.