Selling Into Euphoria: Why Value Investors Must Sometimes Walk Away From Their Biggest Winners

One of the most difficult realities in investing is that success often creates the conditions for future mistakes.

When investors are losing money, caution comes naturally. Risk management feels intuitive. Valuations matter. Balance sheets matter. Business quality matters. When investors are making extraordinary returns, however, discipline becomes much harder to maintain.

Human psychology changes. The same investor who once carefully analysed valuation ratios, competitive advantages, capital allocation policies, and industry structure suddenly begins finding reasons why traditional valuation metrics no longer apply. Narratives become more persuasive than numbers. Expectations become detached from economic reality. Momentum begins masquerading as investment skill.

This phenomenon has repeated itself throughout financial history. The names change. The technology changes. The stories change. Human nature does not.

Over the past several months, we have become increasingly uncomfortable with the valuations assigned to many companies in the semiconductor and broader technology industries. The excitement surrounding artificial intelligence, accelerated computing, data centres, and high-bandwidth memory has produced extraordinary gains for many investors. Some of these gains are entirely justified. Others appear increasingly speculative.

As value investors, our obligation is not to predict how long momentum will continue. Our responsibility is to compare price against intrinsic value and allocate capital where future expected returns appear most attractive.

That distinction is critical.

The market frequently rewards momentum investors during periods of speculative enthusiasm. There is nothing inherently wrong with momentum investing as a strategy. Many successful investors have built exceptional track records using momentum-based approaches. We are simply not among them. Our investment methodology has always been grounded in business fundamentals, valuation discipline, and long-term capital allocation. Therefore, when a stock becomes extremely overvalued relative to our assessment of intrinsic value, selling becomes a methodological necessity rather than an optional decision.

Recently, that principle led us to reduce our position in Micron Technology. After generating a return of approximately 1,485.57%, we sold half of our position at a price of USD 1,060.45 per share. The remaining shares are expected to be liquidated before year-end.

This decision was not based on dissatisfaction with Micron’s management team, technological capabilities, or execution. Rather, it was based on valuation, industry cyclicality, and forward-looking risk-reward considerations.

The Hardest Sell Decisions Are Usually the Most Profitable Ones

Buying is often easier than selling. When purchasing a stock, investors can rely on extensive research, valuation analysis, and business assessments. The decision is generally supported by evidence that the market is underestimating the company’s future prospects. Selling is more complicated. A stock that has appreciated dramatically often continues to rise after investors sell. This creates psychological discomfort. Nobody enjoys watching a former holding continue climbing higher.

Unfortunately, investing is not about maximising emotional satisfaction. It is about maximising long-term risk-adjusted returns.

Every investment decision should be based on future expected returns rather than past performance. The market does not care whether an investor purchased a stock yesterday or ten years ago. The only relevant question is whether current prices offer attractive returns from this point forward. When valuations become detached from fundamentals, expected returns deteriorate even if business performance remains strong. This is a lesson that generations of investors have learned repeatedly.

Cyclical Businesses Are Not Compounders

One of the key factors influencing our decision relates to the distinction between cyclical businesses and long-term compounders. Not all great investments are great businesses. Similarly, not all great businesses become great investments. Micron has been an excellent investment for us. That does not automatically mean it possesses the characteristics of a long-term compounder.

A true compounder typically exhibits several characteristics:

Examples often cited by investors include Microsoft Corporation, Visa Inc., Costco, and S&P Global Inc. These businesses possess economic characteristics that allow value creation to compound over decades.

The semiconductor industry is different. Even the best semiconductor companies operate within a framework influenced by supply cycles, capital expenditure cycles, inventory cycles, customer purchasing cycles, and technological transitions. Periods of shortage are often followed by periods of oversupply. Periods of extraordinary profitability often attract new investment and new capacity. Eventually, increased supply affects pricing dynamics. The cycle repeats.

This does not make semiconductor companies poor businesses. It simply makes them different businesses. Investors who ignore industry structure during periods of optimism often discover that cyclical economics eventually reassert themselves.

History Is Full of Investors Who Believed “This Time Is Different”

Financial history contains a long list of episodes where investors became convinced that traditional valuation disciplines no longer mattered.

Each period possessed unique characteristics. Each period also shared a common psychological pattern. Investors became increasingly confident that extraordinary growth justified extraordinary valuations. In many cases, the underlying businesses remained fundamentally sound. What changed was the price investors were willing to pay. Eventually, valuation and reality converged. Not always immediately. Not always predictably. But eventually.

The challenge is that speculative phases frequently last longer than rational investors expect. This creates pressure to abandon discipline. Value investors are often accused of being too early. Momentum investors are often accused of being too late. Both criticisms contain elements of truth.

The objective is not perfection. The objective is to maintain a repeatable process across multiple market cycles.

Why Investor Sentiment Matters

Recently, we have spent time reviewing investor discussions across various online forums, including Reddit communities focused on technology and semiconductor investing. We do not use social media sentiment as a primary investment tool. However, sentiment can provide useful context. When discussions become dominated by certainty rather than analysis, caution becomes appropriate. When investors begin assuming uninterrupted growth trajectories far into the future, risk often increases. When valuation concerns are dismissed outright, markets may be approaching excessive optimism.

None of these indicators functions as a precise timing tool. Sentiment cannot identify exact market tops. What sentiment can reveal is the emotional temperature of the market. Markets become vulnerable when enthusiasm reaches extremes.

The quote remains popular because it captures an enduring truth about market behaviour. Opportunities often emerge when emotions distort prices. The challenge lies in acting contrary to prevailing sentiment.

Redeploying Capital Is More Important Than Taking Profits

Selling a stock is only half the decision. The other half involves determining where the capital should be allocated next. Many investors focus excessively on selling while giving insufficient attention to redeployment. From our perspective, the purpose of selling an overvalued stock is not simply to lock in gains. The purpose is to improve the portfolio’s future expected return profile.

Accordingly, the proceeds from our Micron sale have largely been redeployed into companies operating within healthcare, financial services, software, and industrial sectors. These businesses currently trade at valuations that appear substantially more attractive than many technology and semiconductor names.

Importantly, we are not suggesting that these sectors are universally cheap. Nor are we suggesting that technology is universally expensive. Markets are more nuanced than such broad generalisations. Opportunities exist at the individual company level. The key is maintaining sufficient flexibility to reallocate capital when relative attractiveness changes.

Without valuation discipline, investors gradually become asset gatherers rather than capital allocators.

The Seductive Danger of Recency Bias

Behavioural finance teaches us that investors frequently extrapolate recent experiences into the future. This tendency is known as recency bias.

• When technology stocks outperform for several years, investors begin assuming continued outperformance.
• When artificial intelligence drives extraordinary earnings growth, investors begin assuming perpetual growth.
• When semiconductor companies generate record profits, investors begin assuming record profits are normal.

History suggests otherwise. Economic cycles continue. Competitive pressures continue. Technological disruption continues. Capital flows continue. The future rarely resembles the most recent twelve months. This reality creates opportunities for disciplined investors willing to think beyond prevailing narratives.

Successful investing often requires distinguishing between a great business and a great investment opportunity. The two are related. They are not identical.

Investing Is Not About Being Right Every Day

One of the misconceptions surrounding portfolio management is that successful investors consistently sell at market tops and buy at market bottoms. Reality is far less elegant. Most successful investors sell too early. Most successful investors buy too early. The goal is not precision. The goal is favourable probability distributions.

• We do not know exactly when semiconductor valuations will normalise.
• We do not know exactly when investor sentiment will shift.
• We do not know exactly how much additional upside remains.

What we do know is that valuation matters over the long term. We know that cyclical industries eventually experience cyclical outcomes. We know that expected returns decline as valuations rise. And we know that preserving discipline during periods of widespread optimism is often more important than predicting short-term price movements.

Final Thoughts

Every investor eventually encounters a position that generates extraordinary returns. The real test comes afterwards. Will success strengthen discipline? Or will success encourage complacency?

For us, selling part of our Micron position was not a prediction about tomorrow’s stock price. It was an expression of our investment philosophy — a philosophy rooted in valuation, capital allocation, and the belief that risk and reward are inseparable.

There may be additional upside ahead. Momentum may continue for an extended period of time. Artificial intelligence investment may exceed current expectations. All of those outcomes are possible. Yet investing is ultimately a game of probabilities rather than certainties. At today’s valuation levels, we believe capital can earn superior long-term returns elsewhere. Time will determine whether that assessment proves correct.

Until then, we remain guided by the same principles that have shaped investors from Benjamin Graham to Charlie Munger and Warren Buffett.