Loss Ratio | The Structure View

Rika Taute
Apr 25
3 min read

In the previous edition, we looked at how a portfolio appears from a capital perspective - combining performance, scale, and volatility into a single view.

It still assumes the portfolio risk sits with us.

Reinsurance reshapes the loss distribution, often in ways that are not visible in standard performance views.

Two portfolios with identical net loss ratios can carry very different retained risk.

Without understanding that structure, we are not seeing the risk we ultimately keep.

The Decision Question

Is our reinsurance structure aligned with the risk we want to retain?

Suggested Chart

This type of view is often used to show the impact of reinsurance - a movement from gross to net in terms of loss ratio and volatility.

It is directionally useful. We can see that reinsurance reduces both the level of losses and their variability.

But it does not show how that change is achieved.

The shift could come from a proportional reduction across all losses, or from removing extreme outcomes entirely. These are fundamentally different risk profiles, yet they can appear similar in this view.

As a result, it shows the outcome, but not the structure that drives it.

To understand what we actually retain, we need to move beyond summary metrics and look at the shape of the loss distribution.

Instead of showing a single movement in loss ratio and volatility, the view below compares the full distribution of gross and net losses. This makes the impact of different reinsurance structures explicit.

Quota Share reduces losses across the entire range, lowering frequency in each band but leaving the overall shape largely unchanged. Excess of Loss operates differently — it removes losses above the attachment point, truncating the tail and eliminating the most extreme outcomes.

This makes it clear not just that risk has reduced, but how it has been reshaped.

This matters because the retained portfolio is no longer exposed to the same drivers of risk. What remains is increasingly driven by attritional performance, while tail risk has been transferred.

What This Makes Visible

· The full shape of the loss distribution, not just summary metrics

· The proportional impact of QS across all loss sizes

· The attachment point and tail removal from XoL

· The shift from tail-driven to attritional-driven risk

· The losses we no longer retain — not just those reduced

· The remaining sources of volatility after reinsurance

· That similar net loss ratios can come from very different underlying structures

Trade-offs & Risks

Ignores cost of reinsurance

The chart shows risk transfer, but not whether it is economically efficient (premium, commission, reinstatements)

Static view of risk

Based on historical or simulated losses - may not reflect changing exposure, limits, or underwriting mix

No view of capital impact

Tail removal is visible, but not how this translates into capital relief or required return

Sensitive to binning and scaling choices

The visual impression of “tail removal” can be overstated or understated depending on how the distribution is grouped.

No view of timing of losses and recoveries

Does not show when losses occur or when recoveries are received, which can materially impact earnings and capital strain.

This view shows what is transferred - not whether it is worth it.

Decision Boundary

The chart becomes decision-grade when it shows not just that risk has reduced - but what is driving the change and what remains.

Use this when….	Don’t use this when…
You need to understand what risk is retained after reinsurance	You are evaluating the cost or profitability of the reinsurance program
Comparing different reinsurance structures (QS vs XoL vs combinations)	You need a view of capital impact or solvency metrics
Explaining the role of attachment points and tail protection	The audience only needs a high-level summary (LR / volatility)
Assessing whether risk is attritional vs tail-driven	The structure includes complex features (reinstatements, aggregates, fac detail) not reflected
Challenging situations where similar net metrics hide different risk profiles	The data is too limited to produce a credible loss distribution

This view shows what is transferred — not whether it is worth it.