Who Carries the Weight: Capital Allocation and Why It Matters

(Part 3 of “Measuring Risk: From VaR to Expected Shortfall”)

How capital gets divided and why the split matters more than most people realize.

By now, we’ve gone from defining what risk is (Part 1) to understanding how it’s measured more honestly through Expected Shortfall (Part 2).

But there’s one question we haven’t answered yet, and it’s the one that really decides who sleeps at night and who gets the 6 a.m. phone call from the regulator:

Who’s responsible for the risk?

Because measuring total risk is only half the battle. The harder part is figuring out how to divide that risk among all the players (which desks, which business units, which assets) and making sure the capital cushion reflects reality.

That’s where capital allocation comes in.

The invisible backbone of banking

Every major bank sits on a delicate balancing act. On one side, it has capital, the financial buffer meant to absorb losses. On the other, it has portfolios full of risky assets generating returns.

Too little capital, and one shock can bring the whole structure down. Too much capital, and profitability shrinks investors start asking awkward questions.

So the key challenge is precision: how much capital does each part of the bank need to carry to reflect the risk it’s taking?

That’s not just a technical question. It’s a philosophical one.

Because underneath it all, capital allocation is about fairness.

*Federal Reserve guidance on capital requirements and supervisory expectations for risk-taking units.*

From total risk to individual contribution

Let’s imagine the bank as a giant dinner party. The total bill (your total risk) is known. Now comes the tricky part: how do you split it among the guests?

You could divide it equally. You could charge people based on what they ordered. Or you could use a formula that accounts for who drank the most wine.

That’s essentially what risk managers do, except the “bill” is measured in millions (or billions) of dollars, and the dinner guests are loan portfolios, trading desks, or business lines.

We call this risk contribution, in other words, how much each component adds to total portfolio risk.

Enter the Euler principle

To make this fair (and mathematically sound) banks often use something called the Euler allocation principle.

It’s based on a simple yet elegant idea: if your risk measure is homogeneous (meaning doubling all exposures doubles the total risk), then you can decompose total risk into the sum of each component’s marginal contribution.

In plain English:

Each portfolio element pays for the risk it actually creates.

It’s like asking, “If I tweak this one position slightly, how much does the total risk move?” That small change (the sensitivity) defines how much capital that position deserves.

*Academic research on Euler-based risk allocation and marginal contributions in portfolio risk.*

Why Expected Shortfall fits perfectly

Here’s where Expected Shortfall (ES) comes back as the unsung hero.

Remember from Part 1 that VaR wasn’t coherent, it sometimes failed the subadditivity test, meaning diversification didn’t always reduce total risk.

That made allocation under VaR shaky. You could end up punishing units for diversifying or worse, assigning capital inconsistently.

Expected Shortfall fixes that. Its coherence guarantees that when you apply the Euler principle, every unit’s contribution adds up neatly to the total.

No contradictions. No accounting gymnastics.

That’s why, in modern risk systems, ES-based capital allocation has become the standard.

An intuitive example

A typical mismatch: where capital is allocated vs. where risk actually comes from, a gap Expected Shortfall helps correct.

Imagine you manage three lending desks:

• Desk A: Consumer credit
• Desk B: Small business loans
• Desk C: Real estate lending

All three contribute to the bank’s overall credit portfolio.

Now, suppose a market shock increases default rates in real estate but barely affects small business lending. With ES-based allocation, more capital automatically shifts to Desk C, the one amplifying total risk.

That’s fairness in motion: dynamic, data-driven, and mathematically defensible.

The hidden benefits of fair allocation

You might think capital allocation is just about regulatory compliance. But it actually changes behavior.

When capital is allocated fairly:

• Managers see the true cost of the risks they’re taking.
• Business units stop gaming the system to look “less risky.”
• Banks can set meaningful performance metrics such as RAROC (Risk-Adjusted Return on Capital).

In short, good allocation creates accountability. It makes risk visible.

And when people can see their impact on total risk, they start managing it better.

A brief detour: RAROC and the language of accountability

RAROC breaks profitability down into revenue, costs, expected losses, and the economic capital absorbed by a business unit.

RAROC (Risk-Adjusted Return on Capital) is what happens when finance meets reality.

It measures how much return you’re earning for every dollar of risk capital allocated to your unit.

If you’re generating $20 million in profit but carrying $200 million in allocated capital, your RAROC is 10%. That number now becomes your performance benchmark, a mirror that reflects both profitability and prudence.

Under Expected Shortfall allocation, RAROC becomes more meaningful because it’s built on consistent, coherent capital weights.

It’s no longer about who shouts loudest in capital meetings. It’s about who genuinely contributes value per unit of risk.

*OECD insights on risk governance, accountability, and capital oversight.*

When the math meets management

Here’s the beauty of it: behind all the calculus, the equations, and the regulatory jargon, the point of capital allocation is profoundly human.

It’s about fairness. It’s about discipline. It’s about ensuring that when risk turns into loss, everyone carries the right share of the weight.

That’s what Expected Shortfall enables, not just sharper math, but better governance.

The limits of the model

Of course, no model is perfect. Even with ES and Euler allocation, some challenges remain:

• Estimating tail losses requires lots of data (and judgment).
• Correlations can shift faster than the models can adjust.
• Extreme events may still behave beyond the assumptions.

But here’s the thing, imperfection doesn’t make a model useless. It makes it human.

These tools aren’t about predicting the future. They’re about preparing for it.

The philosophical takeaway

Capital allocation is the point where mathematics meets morality.

Because once you’ve measured risk, you have to decide who’s responsible for it and that’s as much an ethical decision as it is a financial one.

Expected Shortfall, with all its neat equations, does something quietly profound: it makes that responsibility transparent. It turns abstract uncertainty into something you can see, share, and plan for.

And maybe that’s the ultimate goal of risk management, not to eliminate risk, but to distribute it wisely.

Wrapping up the series

We started with a simple question. How much risk are we really taking?

Then we saw how Value-at-Risk gave an answer, but an incomplete one. We met Expected Shortfall, a more honest, coherent way to measure risk. And finally, we arrived here, at capital allocation, where numbers meet accountability.

Together, these ideas form a quiet philosophy of modern risk management: Measure clearly. Allocate fairly. Prepare humbly.

Because risk isn’t just math on a balance sheet. It’s the heartbeat of every decision we make kind of like the silent partner in every gain and every loss.

And the models, for all their flaws, are just our way of keeping time.

Continue the series

• Part 1 (When “How Much?” Isn’t Enough: Rethinking Risk Measures)
• Part 2 (How Expected Shortfall Changed the Game)