The Probability Portfolio

Defense spending is at its highest point in decades, and the companies capturing those dollars have changed. A new class of contractors is winning Pentagon business with AI-driven systems, satellite infrastructure, and advanced aerospace technology. This free research report profiles five of them. You'll find what each company does, why it's winning contracts, and what the growth case looks like from an investor's perspective. These aren't household names yet. That's the point. Download the free report and see why analysts are paying attention to this corner of the market before the rest of Wall Street catches on.

Download the free report

Welcome to The Edge. If you are reading this, you are part of a small group of subscribers I assume have moved past the basics of personal finance. You are not here for general advice about emergency funds and diversified index funds, though those things matter and I would not dismiss them. You are here because you want to think more rigorously and more systematically about how capital actually compounds over time and what, specifically, separates investors whose portfolios drift and erode from those whose compound with increasing reliability.

Today I want to go deep on something that institutional investors understand and apply daily but that almost never appears in mainstream financial education: probabilistic portfolio construction. This is not a niche approach. It is how professional capital allocators think about every position, every sizing decision, and every rebalancing action. And it is available to individual investors who are willing to do the intellectual work.

The payoff for that work is significant. Investors who apply even a basic probabilistic framework to their decision-making tend to take better risks, size positions more intelligently, and recover from market dislocations faster and more deliberately than those operating on conviction and intuition alone.

The Problem with How Most People Invest

Most individual investors make decisions based on one of three inputs: what has worked recently, what a financial advisor or media personality endorsed, or what aligns with their current narrative about where the economy is heading. These are not irrational inputs. But they share a common flaw: they treat the future as a single outcome rather than a distribution of possible outcomes.

Conviction-based investing is the natural result of this single-outcome thinking. You believe a sector, company, or thesis is correct. You allocate heavily based on that belief. Sometimes the conviction is right and you look prescient. Sometimes it is wrong and you experience a painful loss that you rationalize as bad luck rather than a structural flaw in your process.

The problem runs deeper than individual wins and losses. Research in behavioral finance consistently documents a specific finding that is worth sitting with: investor confidence and investor accuracy are not just uncorrelated. In many documented cases, they are inversely correlated. The most confident investors tend to be most confidently wrong at the worst possible moments, precisely because high conviction tends to peak at the same time as asset valuations. You become most certain about a position when the evidence in your favor has been accumulating for a while, which is also when the expected future return of that position is lowest.

Probabilistic investing starts from an entirely different premise. Instead of asking what you believe will happen, it asks: what is the realistic distribution of outcomes here, how confident am I in that distribution, and how do I build a portfolio that performs reasonably well across most of that distribution rather than betting everything on one scenario being correct?

Expected Value: The Foundation

Every probabilistic investment decision begins with expected value. Not what you think will happen. The probability-weighted average of everything that realistically could happen.

That nine percent expected value looks reasonable on paper. But here is where it gets interesting. Compare that position to one with the following distribution: a seventy percent chance of returning twelve percent and a thirty percent chance of returning two percent. The expected value is 0.7 x 12 plus 0.3 x 2, which equals 9.0 percent. Identical expected values. Dramatically different risk profiles.

The first position has wider variance and a meaningful probability of a fifteen percent loss. The second position has tighter variance and no scenario with a negative return. For most portfolios, the second position is preferable at the same expected value because the lower variance means your compounding is less disrupted by drawdowns.

Which one you choose depends on your existing portfolio composition, your correlation exposure, your liquidity needs, and your genuine capacity for drawdown, not on which thesis you find more intellectually compelling. The expected value framework forces you to have that conversation explicitly rather than implicitly.

The Distribution, Not the Point Estimate

One of the most practically useful shifts you can make as an investor is replacing point estimates with probability distributions in your investment analysis.

A point estimate is: ‘I think this company’s stock will appreciate thirty percent over the next twelve months.’ A probability distribution is: ‘I think there is a fifteen percent chance this stock appreciates more than fifty percent, a forty percent chance it appreciates between ten and fifty percent, a thirty percent chance it returns between negative ten and positive ten percent, and a fifteen percent chance it declines more than ten percent.’

The distribution version contains dramatically more honest information. It forces you to think about what would have to be true for each outcome, which often reveals that the upside scenario depends on assumptions that are less certain than you initially assumed. It also forces you to quantify the downside scenarios rather than mentally categorizing them as ‘low probability’ without actually assigning a number to them.

You do not need to be a quant to do this. You need to be honest and disciplined about explicitly considering multiple scenarios rather than anchoring on the most favorable one. The discipline of writing down the distribution, even roughly, is worth more than the precision of the numbers.

The Correlation Trap

Most investors who believe they are diversified are not. This is one of the most persistent and costly misconceptions in personal portfolio management. Diversification is not about owning assets that sound different from each other. It is about owning assets whose returns are genuinely uncorrelated, meaning that when one falls sharply, the others do not fall with it for the same underlying reasons.

The 2022 market environment provided a brutal illustration of correlation failure. Investors who believed they had a balanced allocation between equities and investment-grade bonds discovered that rising interest rates created sharp losses in both simultaneously. The traditional sixty-forty portfolio had one of its worst years in modern history, not because the theory of diversification failed, but because equities and bonds in that specific environment shared a common return driver: sensitivity to interest rate changes.

True diversification requires thinking in terms of return drivers rather than asset class labels. Return drivers are the underlying economic factors that cause an asset to go up or down. The relevant drivers for most portfolios include economic growth sensitivity, interest rate sensitivity, inflation sensitivity, credit risk exposure, liquidity risk, and currency exposure.

A portfolio that holds large-cap US equities, growth-oriented tech positions, corporate bonds, and real estate investment trusts might appear diversified across four asset classes. But if all four of those positions are significantly sensitive to interest rate changes, the portfolio is less diversified than it appears. Understanding the return driver overlap in your portfolio is more important than counting the number of different ticker symbols you hold.

Position Sizing as a Probabilistic Exercise

The Kelly Criterion is a mathematically derived formula for optimal bet sizing that originated in information theory and has found wide application in professional investing, poker, and other high-stakes decision environments. The core insight is that there is a theoretically optimal amount to stake on any given opportunity, one that maximizes long-term portfolio growth without risking the ruin that comes from over-betting.

Most sophisticated investors use a fraction of the full Kelly allocation, typically between one quarter and one half, to account for the estimation error inherent in the probability and payoff inputs. You are not omniscient about the distribution, so you should not bet as though you are. Even at half Kelly, the position sizing discipline tends to produce significantly better long-run results than equal-weight or conviction-weight allocation.

The practical discipline the Kelly framework imposes is this: every position requires an explicit expected win probability and an explicit expected win-to-loss ratio. If you cannot produce even rough estimates of those two numbers, you are not investing probabilistically. You are speculating on a feeling. There is nothing wrong with acknowledging that distinction, but it is important to make it consciously rather than operating on conviction while believing you are being systematic.

Building a Scenario Framework

Here is a practical approach to applying probabilistic thinking to your portfolio construction without requiring a quantitative finance background.

Build three explicit scenarios for the next eighteen to twenty-four months. Assign a probability to each. Ensure they sum to one hundred percent. Then evaluate each current position and each prospective investment against all three scenarios.

Calculate the probability-weighted expected return of your full portfolio across all three scenarios. If the weighted return is only attractive in the bull case scenario, you are implicitly betting on one outcome. That may be intentional. But it should be a conscious choice with explicit risk acknowledgment, not an accidental concentration that emerges from uncritical position accumulation.

Most investors who run this exercise for the first time are surprised to discover how concentrated their portfolios are in a single scenario. The exercise creates the visibility that makes deliberate risk management possible.

Rebalancing as a Systematic Edge

Rebalancing is typically described in risk management terms, as a way to prevent your portfolio from drifting too far from your intended allocation as individual positions appreciate or decline. This framing is accurate but incomplete.

Rebalancing is also a return-generating discipline. When you systematically sell assets that have appreciated beyond their target allocation, you are selling things whose expected future returns are relatively lower because valuations have expanded. When you buy assets that have declined below their target, you are buying things whose expected future returns are relatively higher. Done consistently and systematically, this is a mechanical implementation of the buy low, sell high principle that almost no one actually executes with discipline in practice.

The research on rebalancing frequency suggests that threshold-based rebalancing, where you rebalance whenever any position drifts more than a defined percentage from target, tends to produce slightly better risk-adjusted outcomes than rigid calendar-based rebalancing because it is more responsive to actual market movements. The specific threshold matters less than the consistency of application. Pick a rule and follow it without exception, because the value of rebalancing comes almost entirely from its disciplined execution rather than its precise timing.

The critical failure mode with rebalancing is emotional override. The moments that most require rebalancing, when a category has declined sharply and needs buying or when a winner has run and needs trimming, are precisely the moments when it feels most psychologically difficult to execute. Building the rule in advance and committing to it eliminates the emotional override from the equation.

Tools for the Individual Probabilistic Investor

The institutional infrastructure for probabilistic portfolio management is increasingly accessible to individual investors. Galaxy.ai allows me to quickly analyze earnings reports, macro research, and sector data in compressed, decision-relevant formats. The ability to move from raw information to probability-adjusted insight in minutes rather than hours is a meaningful edge that simply did not exist for individual investors five years ago.

Most brokerage platforms now offer correlation matrices and factor exposure tools within their analytics suites. These are worth spending serious time with. Understanding which of your positions share meaningful return driver overlap is a critical input to building a genuinely diversified portfolio rather than one that just looks diverse on the surface.

A structured decision journal is one of the highest-leverage tools available and it costs nothing. For every significant position you take, record the probability distribution you assigned, the expected value calculation, the scenarios you evaluated, and the specific reasoning behind the sizing decision. Review it quarterly. The feedback loop between your documented probability estimates and the actual outcomes is how you calibrate and improve your investment judgment over time. It is also the clearest window into your own cognitive biases.

For staying current on developments that could shift your probability estimates, Beehiiv.com hosts some of the best independent research newsletters in finance, several of which provide genuinely insightful probabilistic analysis of current market conditions. Building a curated reading stack from these sources is a high-return investment in the quality of your input information.

The Honest Caveat

Probabilistic portfolio construction does not guarantee outperformance. Nothing does, and anyone telling you otherwise is selling something. What it does guarantee is a more rigorous and improvable process. You will make some probabilistic calls that are wrong. Markets are complex, the future is genuinely uncertain, and humility about the limits of any analytical framework is appropriate.

What the probabilistic approach does is reduce the magnitude of the errors that come from overconfidence, and those errors tend to be the catastrophic ones. The investor who bets ninety percent of their portfolio on a single conviction thesis and is wrong does not recover easily. The investor who sizes every position based on an explicit expected value calculation and a defined probability distribution takes losses that are painful but survivable and instructive.

Over long time horizons, better decision-making processes produce better outcomes with a reliability that short-term noise obscures completely. The investor applying a systematic framework will not always outperform in any given year. But averaged across years, across market cycles, and across the full range of outcomes that a long investing lifetime will produce, the systematic approach compounds more effectively than the conviction approach. Not dramatically. Not every time. But consistently.

That consistency is the edge. It is available to any investor willing to do the intellectual work of building it.

This Week’s Action Step

Alex Rivera

Wealth Architect, The Wealth Grid

wealthgridhq.com