From a number to a regime
Earlier in this rung you built up enterprise risk management from the inside out: you put every risk on one balance sheet, chose a risk appetite, measured the danger with a risk measure, and worked out how much economic capital should stand behind the promises. That was the firm answering, for itself, *how much capital is prudent?* This guide flips the seat. Now a regulator — answerable to the public, not the shareholders — asks the same question and turns the answer into law. The result is a capital regime: a binding rulebook that says how to value the balance sheet, how much capital to hold, and what happens if you fall short.
The most influential of these regimes is Solvency II — the European Union's comprehensive, risk-based supervisory framework, in force since 2016. Before it, many countries supervised insurers with a thin rulebook of simple ratios: hold capital equal to some fixed percentage of premiums or reserves, regardless of how risky the business actually was. The Solvency II framework threw out that crude approach. Its founding idea is that the capital an insurer must hold should genuinely reflect the risks it actually runs — a catastrophe writer and a plain term-life shop should not be measured by the same blunt ratio. That single shift, from fixed ratios to risk-based capital, is what makes the whole regime worth a guide of its own.
Three pillars: capital is never enough on its own
Solvency II stands on three pillars, and the deepest lesson of the whole regime is hidden in that number. A safe insurer needs more than a big pile of capital; it needs to be well run and to be transparent about what it is doing. Pillar 1 is the quantitative pillar — the numbers: how to value the balance sheet, and how much capital to hold. Pillar 2 is the governance and supervision pillar — the qualitative side: a sound risk-management system, fit-and-proper people, and the firm's own forward-looking self-assessment. Pillar 3 is the disclosure pillar — public and supervisory reporting, so the market and the regulator can actually see the risks.
Pillar 2 is where actuarial work meets the boardroom most directly, through the Own Risk and Solvency Assessment, or ORSA. The standard capital formula is a one-size health check given to every insurer at the door; the ORSA is the firm doing its own thorough check-up. It is the company's own forward-looking judgement of *all* its material risks and whether its capital will stay adequate over the whole business plan — several years out, not just the next twelve months — explicitly tied back to the board's risk appetite. It is not a single regulator-defined number; it is a process the firm owns and must actually use to make decisions. A glossy ORSA report that nobody read before deciding anything has missed its entire point.
Two lines on the wall: the SCR and the MCR
Inside Pillar 1 the regime first insists on a market-consistent balance sheet: assets and liabilities are both valued at realistic, current-market terms. The liability — the technical provision — is the best-estimate present value of future claims plus a risk margin on top, the extra cushion that a buyer would demand to take the liabilities off your hands. That risk margin is computed by the cost-of-capital method you met earlier in this rung. On top of that liability sit two capital requirements, like two warning lines painted on the wall of a reservoir.
The higher, comfortable line is the Solvency Capital Requirement (SCR). It is calibrated so that, over the coming year, the chance of becoming insolvent is no more than 0.5% — the capital needed to survive a 1-in-200-year loss. In the language of the previous guides, the SCR is a one-year, 99.5% Value at Risk on the change in the insurer's own funds. The lower, emergency line near the bottom is the Minimum Capital Requirement (MCR) — a simpler calculation, bounded roughly between 25% and 45% of the SCR (with an absolute monetary floor). The pairing is the whole point: the SCR is the level you are meant to manage to, and the MCR is the point of no return.
The consequences are graded, and that grading is what makes two lines better than one. Breach the SCR and the supervisor does not shut you down — it demands a credible recovery plan to restore capital within months. Breach the MCR and the ultimate action follows: the authorisation to operate can be withdrawn. A common confusion is to think an insurer must always sit far above the SCR. It need not. A firm can dip below the SCR in a genuinely bad year and survive through an orderly recovery; it is sliding toward the MCR that signals the true endgame. The SCR is the early-warning trigger; the MCR is the final backstop.
Own funds (assets - technical provisions): 120 SCR (1-in-200, 99.5% VaR over 1 year): 80 -> ratio 120/80 = 150% OK MCR (here 35% of SCR = 0.35 * 80): 28 -> ratio 120/28 = 429% OK Bad year: own funds fall to 70. 70 < SCR(80) -> below SCR: recovery plan required (still trading) 70 > MCR(28) -> above MCR: licence not yet at risk
The standard formula or your own model
How is the SCR actually computed? Solvency II offers two routes, and the choice is real. The first is the standard formula — the regulator's prescribed recipe. It is a fixed menu of risk modules (market, life, non-life, health, counterparty default, operational), each stressed by given parameters, then combined through prescribed correlation matrices that grant a set diversification benefit — because a firm rarely suffers every shock at once. The standard formula is calibrated for an 'average' European insurer: simpler, comparable across firms, and cheap to run.
The second route is an internal model — a firm's own bespoke model of its risks, often a full stochastic simulation that draws thousands of possible futures and reads the SCR off the resulting distribution of own-fund losses. This is the regime's gateway to the firm's own economic-capital view, now blessed for regulatory use. But it must be approved by the supervisor and pass strict tests — above all the use test: proof that the model genuinely drives how the business is run, not merely produces a smaller capital number. Firms can also use a *partial* internal model, bespoke where they differ from the average and standard elsewhere.
Why bother building your own model? Because the standard formula can badly mis-state capital for a firm whose risk profile departs from the average — a monoline catastrophe writer, or a life office with unusual guarantees. A well-built internal model can produce a fairer SCR (sometimes lower, sometimes higher) while sharpening management's grip on its real risks. The honest caveats are real: internal models are expensive to build, validate and maintain; approval is demanding; and there is a permanent temptation to tune assumptions toward a flattering figure. That temptation is precisely why the use test and ongoing supervisory scrutiny exist — and it is a useful reminder that a model is a map of reality, never reality itself.
Switzerland's road: the Swiss Solvency Test
Switzerland faced the very same question — how much capital makes an insurer safe? — but answered it a little differently, and a few years earlier. The Swiss Solvency Test (SST), introduced by the regulator FINMA in 2006, is a market-consistent, principles-based regime whose spirit is close to Solvency II's: it values assets and liabilities at market-consistent terms and adds a risk margin. What makes the SST such an illuminating companion is one deliberate difference in *how it measures the danger*.
Here is the heart of it. Where Solvency II's SCR is a 99.5% Value at Risk — the loss at the 1-in-200 *cut-off*, a single point — the SST uses a 99% Tail Value at Risk, also called expected shortfall: the *average* loss across the worst 1% of outcomes. You met this distinction earlier in the rung, and Solvency II versus the SST is its real-world showcase. VaR tells you where the bad tail begins; Tail VaR tells you how bad it gets once you are in there. A regime built on VaR can be blind to a catastrophe lurking just beyond the cut-off; a regime built on Tail VaR deliberately looks at the whole tail, so a fatter, more dangerous tail demands more capital.
The lesson is bigger than the two regimes. Two careful regulators, both aiming at roughly the same level of safety, chose different risk measures — and that choice is not academic. It is a genuine policy decision about how seriously to weight catastrophe, and it is why headline solvency ratios from the two regimes cannot be compared at face value: the rulers differ. This is also why the SST leans heavily on explicit, severe predefined scenarios sitting alongside its modelled distribution, baking stress and scenario thinking into the regime rather than bolting it on. Different countries, different dials — but the same honest question underneath: when the worst really happens, will the promises still be paid?
Where the actuary stands in all this
Step back and see how much of this whole ladder lands here. Valuing the technical provisions market-consistently uses the discounting and present-value craft from the interest and ALM rungs. The risk margin uses the cost-of-capital idea. The SCR is the firm-wide economic capital question made into law, expressed in the risk measures of the previous guides. The internal model is the stochastic simulation of loss distributions you have been building toward all along. Solvency is not a separate subject bolted onto actuarial science — it is where almost every thread you have learned is pulled together onto one balance sheet.
It is also where the actuary's honesty is tested hardest, because the regime hands you levers that move the capital number — the discount curve, the stresses, the model assumptions — and a flattering choice on any of them can make a fragile firm look robust. A capital regime is only as truthful as the judgement poured into it. The deepest takeaway of this rung is the same one that opened it: every promise the firm has made sits on one balance sheet, and a finite amount of real capital must stand behind it. Solvency II and the SST are two careful, imperfect attempts to make sure that capital is genuinely there — measured, governed, and disclosed — on the day the worst arrives.