Explore three distinct risk measurement frameworks—Asset-Only, Liability-Relative, and Goals-Based—and discover how each transforms portfolio construction and performance evaluation.
Well, let’s be honest here: sometimes the biggest challenge in portfolio management isn’t choosing the next big stock or identifying which bond to buy. It’s all about figuring out how to define and measure risk. And guess what? There’s actually more than one way to look at risk. Maybe you’re a big institution trying to outperform a market benchmark, or perhaps you’re a pension fund that wants to ensure it can pay retirement benefits. Heck, maybe you’re just managing your personal wealth and have a few non-negotiable life goals, like your child’s college tuition or that dream house by the ocean.
In this section, we’ll compare three core risk perspectives in asset allocation:
• Asset-Only: Focused on portfolio volatility, Sharpe ratios, and similar metrics.
• Liability-Relative: Centered on matching or exceeding liabilities to minimize a surplus gap.
• Goals-Based: Organized around personal or institutional goals to manage the probability of achieving each one.
We’ll unpack each approach, check out their key risk measures, and explore how these loads of theory meet real-world portfolios. Even if you’ve been working in finance for a while, you might pick up a fresh perspective on risk—and if you’re new to all this, don’t worry. By the end of this discussion, you’ll be able to explain the difference to your best friend at the coffee shop, or at least to a colleague in the next cubicle.
Let’s start with a quick review. Risk in investment terms is often about uncertainty—uncertainty in returns, in achieving objectives, or in meeting obligations. The challenge is: whose objectives or obligations are we talking about?
• Asset-Only. Typically about beating a benchmark or controlling volatility.
• Liability-Relative. Specifically about covering liabilities, often for pensions or insurance.
• Goals-Based. For individuals (or organizations) who have specific targets—like college funds, philanthropic gifts, lump-sum outlays, or retirement income.
Each perspective offers a unique “lens” through which portfolio managers can see the world. And, in practice, large institutions can have multiple lenses on at once—like sporting a pair of fancy glasses with different filters for their various portfolios.
Below, let’s dive deeper into each approach.
Under the Asset-Only (AO) approach, a portfolio is often measured and judged based on its returns relative to broad market indices or absolute-return targets. The primary risk metrics might include:
• Volatility (standard deviation of returns).
• Sharpe ratio (excess return per unit of risk).
• Value at Risk (VaR), reflecting potential maximum loss over a particular time horizon.
• Max drawdown (for folks who really want to know how much they can lose in a worst-case scenario).
When you hear “Asset-Only,” think: “How is this portfolio doing in and of itself—without factoring in external obligations?” That might be perfectly fine for, say, an endowment that wants to outperform a 70/30 benchmark or an individual who just wants to maximize returns for a given risk tolerance.
The AO approach is straightforward: you assess risk by looking at the portfolio’s own characteristics. This makes it relatively simple to communicate: “We lost only 10% in a bear market while the benchmark fell 15%—hooray.” You watch your volatility, maybe track your performance vs. a recognized index, and call it a day.
But what if you’re actually investing to meet liabilities—like monthly pension payments—rather than to outperform a generic stock index? In that case, the AO approach might miss the mark because it doesn’t explicitly incorporate your liabilities. You might be “winning” relative to some index while still being unable to pay out the obligations you owe.
Remember a time you had a family member brag about beating the market? That’s the essence of asset-only thinking. Let’s say you have a portfolio with 50% in equities, 40% in bonds, 10% in alternatives. You measure success by alpha over a stated benchmark, say the S&P 500 for equities and a broad bond index for the fixed-income portion. As you rebalance, you’re mainly aiming to keep portfolio risk (as measured by standard deviation) within a set band. Perfectly fine for many investors. But for a pension plan? Possibly incomplete, as we’ll see next.
If you work for a pension fund or an insurance company, or any entity that must meet specific payments in the future, the Liability-Relative (LR) approach is your bread and butter. Here, the “risk” is not about absolute volatility of your asset portfolio alone. Instead, it’s about your surplus volatility—i.e., how much the difference between your assets and your liabilities bounces around over time.
Mathematically, you might define the Surplus (S) as:
S = A − L,
where A = Value of Assets, L = Present Value (or Market Value) of Liabilities.
The risk perspective shifts from “I hope I don’t lose money in the market” to “I hope I always keep enough to pay my obligations.” Volatility is measured in terms of how that surplus behaves.
A hallmark of liability-relative management is to match the duration or some key characteristics of the portfolio to that of the liabilities. For example:
• If you know your liabilities have an effective duration of 10 years (like a stream of pension payments), investing in a fixed-income portfolio with a similar duration can reduce mismatch risk.
• Asset-Liability Management (ALM) models can help you see how changes in interest rates, inflation, or other factors impact both assets and liabilities.
The big plus is that the LR approach directly addresses an organization’s primary need: meeting liabilities. This is particularly relevant for defined-benefit pension plans. By focusing on surplus volatility, managers can hone in on techniques to reduce the mismatch in timing and magnitude of cash flows.
The downside? Well, focusing purely on matching liabilities might come at the expense of growth opportunities. Sometimes the best solution for meeting future obligations is investing in higher-return (and higher-risk) assets. But that can increase the volatility of your surplus. It’s about striking a balance—kind of like wearing a seatbelt: you might limit some range of movement, but you’re also safer if you crash.
Imagine a pension plan that invests heavily in long-duration bonds because its retirees, on average, are 20 years away from hitting peak retirement benefit payouts. By aligning the bond portfolio’s duration to match the liability stream, the plan’s managers reduce the risk that interest rate fluctuations will cause a mismatch between assets and liabilities. If interest rates rise, the market value of their liabilities goes down (because pension liabilities are often discounted at the relevant market interest rate), but so does the value of their bonds—however, ideally at a comparable rate. The manager’s job is to keep that surplus line stable, or at least stable enough to cover the promised benefits.
The Goals-Based (GB) approach is quite personal—though institutions can adopt it, too. It effectively breaks a portfolio into “goal buckets.” Each bucket has its own objective (like a future college payment or a philanthropic bequest), time horizon, and risk tolerance. Instead of focusing on a single volatility number or a single surplus measure, you ask: “What is the probability of achieving each goal? And how do we handle shortfalls if we miss the mark?”
One of the most important risk metrics in this framework is shortfall probability, or the chance that a particular goal won’t be reached. You’ll often see this expressed in simulation-based terms. For instance, Monte Carlo or scenario analysis might measure how likely you are to accumulate, say, $200,000 by your child’s 18th birthday.
Sometimes, you might care less about the magnitude of returns and more about the binary success/failure of achieving that goal. You set a threshold (“I need at least X dollars by Year Y”), and you track the probability of crossing that threshold. This is a very different perspective than the typical “invest and measure portfolio volatility,” but it’s arguably more intuitive for individuals and smaller institutions with distinct, non-fungible goals.
Goals-based investing is super intuitive. You might say, “I have a big mortgage balloon payment in 5 years, and I can’t risk not having enough in 5 years.” So, you create a bucket with a conservative asset allocation. Meanwhile, another bucket might be for a dream vacation in 15 years, so you can be more aggressive there. You measure risk in terms of each bucket’s shortfall probability, not in terms of the entire portfolio’s standard deviation.
However, dividing the portfolio into smaller sub-portfolios can create inefficiency. You might lose the benefit of a single, diversified pool. Also, some advisors argue it’s more rational to manage your entire net worth as a single portfolio since money is (in principle) fungible. But from a behavioral perspective, we humans often prefer to designate “buckets” with different risk profiles that map onto our actual life goals.
I recall meeting a friend who said something like, “I have my ‘vacation condo in the mountains’ fund, and I don’t really care if it swings around in value for the next 10 years, because that’s my dream for retirement.” But for their kid’s college fund, it’s a different story. They cannot risk significant losses as the child approaches 18. So, they keep that bucket in a more conservative set of assets. That’s goals-based investing in a nutshell.
So, how do these differ if we lay them out side by side? Let’s try a simple diagram to visualize the distinctions in risk focus.
graph TB
A["Asset-Only"] --> B["Focus on Portfolio Volatility <br/> and Market Benchmarks"]
A --> C["Risk Metric: Standard Deviation, VaR, etc."]
D["Liability-Relative"] --> E["Focus on Surplus Volatility <br/> (Assets − Liabilities)"]
D --> F["Risk Metric: Surplus Risk, ALM"]
G["Goals-Based"] --> H["Focus on Probability <br/> of Achieving Specific Goals"]
G --> I["Risk Metric: Shortfall Probability <br/> per Goal Bucket"]
Each approach targets a different aspect of risk:
• Asset-Only: “Is my portfolio stable or beating the market?”
• Liability-Relative: “Can I cover my obligations?”
• Goals-Based: “Am I meeting each of my distinct goals?”
• Standard deviation, Sharpe ratio, Value at Risk.
• Benchmark-centric performance metrics.
• Surplus volatility, funded ratio (assets ÷ liabilities).
• Key rate duration matching and asset-liability matching (ALM) models.
• Monte Carlo simulations for each goal.
• Shortfall probability for each goal bucket.
• Stress testing for different scenarios.
It’s not uncommon for large organizations to deploy multiple approaches. For instance, a large university might have:
• A pension pool that follows a liability-relative approach (to ensure retiree benefits).
• An endowment that’s managed more like an asset-only approach (aiming to maximize return for scholarships, capital projects, etc.).
• Specific donation-based “chairs” or philanthropic grants that might adopt a goals-based perspective, ensuring certain projects get funded with high probability.
Similarly, a high-net-worth individual could manage her overall portfolio by combining these frameworks: an inheritance trust (liability-like promises to heirs), a philanthropic kitty (goals-based, with a target donation amount), and a personal investment account (asset-only, for personal net worth).
• Mixing risk concepts unintentionally. For instance, adopting an asset-only measure for a plan that must meet liabilities can lead to underfunding.
• Over-optimizing for a single measure. Chasing minimal surplus volatility in a pension plan may lead to insufficient returns for future obligations.
• Bucket isolation. Goals-based investing helps psychologically, but siloing assets too strictly can miss out on overall portfolio diversification benefits.
• Lack of dynamic review. Markets change, liabilities shift, and personal goals evolve. No matter the approach, risk measurement must be an ongoing process.
For the Level III CFA exam, understanding these frameworks is crucial. Scenario-based questions might ask you to recommend a risk measure tailored to an institutional investor with specific liabilities or to propose a goals-based plan for a private wealth client with multiple life objectives. Keep these tips in mind:
• Precisely define the investor’s objective. Is it about meeting retirement payouts (liability-relative) or maximizing returns within a risk tolerance (asset-only) or achieving a set of goals (goals-based)?
• Propose risk metrics consistent with the chosen framework (e.g., standard deviation for asset-only, surplus volatility for liability-relative, shortfall probability for goals-based).
• Provide a rationale for why a specific framework is appropriate (“The pension has a legal obligation to meet retiree benefits, thus surplus volatility is the priority”).
• Where needed, highlight possible trade-offs (growth vs. risk management, or synergy among multiple frameworks).
Consider ACME Corporation, which has a large defined-benefit pension plan and a separate endowment fund for philanthropic scholarships. Meanwhile, Alexis (ACME’s CFO) also manages her personal wealth with multiple goals—such as early retirement, a second home, and philanthropic giving.
• ACME’s Pension Plan → Liability-Relative Approach. The plan invests in long-duration bonds to match the liability duration of retirees’ future benefits. It measures success via surplus volatility: the difference between plan assets and plan liabilities.
• ACME’s Endowment → Asset-Only Approach. The endowment invests in a diversified global portfolio with a target to exceed a blended equity-bond benchmark. Risk is measured by standard deviation of returns and tracking error versus the benchmark.
• Alexis’s Personal Wealth → Goals-Based Approach. She sets up distinct “goal buckets:” (1) Early Retirement Fund, (2) Kid’s College Tuition, and (3) Dream Beach House. Each goal has a different time horizon and risk tolerance. She conducts Monte Carlo analysis to estimate the shortfall probability for each goal.
In each segment, a different risk concept shapes portfolio construction, monitoring, and rebalancing strategies.
So, which risk concept is “best”? That depends on what you’re trying to accomplish, how you define success, and what constraints you face. Asset-only is by far the simplest if benchmarks and absolute or relative return metrics are all that matters. If you’re an entity with substantial liabilities, you had better track that surplus to avoid underfunding. And if you’re an individual or institution with multiple specific, non-negotiable targets, goals-based can bring clarity and peace of mind—even if it leads to a somewhat “bucketed” approach.
Every approach has its place. Often, real-world investors adopt a blend, carefully describing the portion of the portfolio tied to liabilities or goals and the portion that is free for more growth-oriented, asset-only investing. By keeping these frameworks in mind, you’ll be equipped to structure a portfolio’s risk management in a way that best serves the investor—be it a multinational corporation, a charitable foundation, or, well, you personally trying to retire at the beach.
• Fabozzi, Frank J., “Handbook of Finance, Volume 2: Investment Management and Financial Management.”
• Leibowitz, Martin L., “Pension Fund Excellence: Creating Value for Stockholders.”
• CFA Institute, “Asset Allocation Readings on Risk Concepts” in the CFA Program Curriculum.
For deeper insight, also revisit Section 3.1 in this volume on Effective Investment Governance (especially if you’re dealing with multiple stakeholders and multi-portfolio structures). You might also later look at Sections 3.7 and 3.10, which address how constraints and rebalancing interplay with your chosen risk framework.
Important Notice: FinancialAnalystGuide.com provides supplemental CFA study materials, including mock exams, sample exam questions, and other practice resources to aid your exam preparation. These resources are not affiliated with or endorsed by the CFA Institute. CFA® and Chartered Financial Analyst® are registered trademarks owned exclusively by CFA Institute. Our content is independent, and we do not guarantee exam success. CFA Institute does not endorse, promote, or warrant the accuracy or quality of our products.