Claim-Demand Inflation: Why Demand Letters Need Their Own Severity Dataset

Claim-demand inflation is the rate at which plaintiff settlement demands are rising, and it is consistently outpacing the severity assumptions embedded in casualty reserving and pricing models. When demand letters are extracted, structured, and tracked as their own dataset, reinsurers gain a forward-looking severity signal that reveals inflation months before settled-claim data catches up.

Why does claim-demand inflation need to be measured separately?

Claim-demand inflation needs to be measured separately because the settled-claim data that drives traditional severity analysis is a lagging indicator. A demand letter arrives today and states the plaintiff's opening position; the claim settles eighteen months later at a negotiated figure that may be higher or lower than the demand. By the time the settlement enters the reserving triangle, the demand that signaled the severity trend is two years old, and the triangle still reflects a severity environment that no longer exists.

The gap between demand inflation and settled-claim inflation is the blind spot that long-tail reserving models are slowest to close. In a general liability portfolio, the average demand on new claims may have risen by twelve percent in a year while the settled-claim severity used in the reserving triangle has risen by six percent. The six-percent figure is what the actuary projects forward; the twelve-percent figure is what the portfolio is actually experiencing. The reserving committee makes its decision on the six-percent figure and is surprised by the twelve-percent reality when it eventually shows up in the triangle, usually as adverse development.

For casualty reinsurers, this gap directly affects pricing. A treaty pricing model built on settled-claim severity is pricing yesterday's inflation rate into tomorrow's treaty. A pricing model that also incorporates demand-letter data sees the inflation that is already in the pipeline but not yet in the settlements, and prices accordingly. The difference is not academic; it is the margin between a treaty that performs as priced and one that develops adversely from the moment it is bound.

What goes wrong when demand-letter data is not systematically captured?

Demand-letter data not systematically captured fails in five ways: the inflation signal in demands is lost because demands are buried in claim files, severity trends are measured from settlements that are years old, reserve adequacy erodes silently as demands outpace reserves, jurisdiction-level inflation differences are invisible, and pricing actuaries build loss picks from an inflation rate that is already obsolete.

The demand letter is the single richest severity signal a claim produces. Not capturing it systematically is a data-architecture failure that costs casualty reinsurers in reserving accuracy and pricing precision.

1. How is the demand-letter inflation signal lost in claim files?

The demand-letter inflation signal is lost because demand letters are stored as documents in individual claim files, not extracted as structured data. A portfolio of ten thousand claims may contain five thousand demand letters, each with a demanded amount, injury allegations, and jurisdiction data that collectively describe the severity trend the portfolio is experiencing. But because that data is locked in unstructured documents, it is invisible to the actuarial analysis.

Extracting demand-letter data at scale, using document processing technology that reads the demand, pulls the key fields, and structures them into a dataset, is what converts the demand-letter archive from a filing exercise into a severity intelligence asset. Once structured, the data can be segmented, trended, and compared to reserving assumptions in near real time.

2. Why does settled-claim severity data lag the market?

Settled-claim severity data lags the market because the average casualty claim takes eighteen to thirty-six months from demand to settlement. The settlement that enters the reserving triangle this quarter reflects the demand environment of two years ago. The demand environment of today will not appear in settled-claim data until two years from now.

This lag is structural, not fixable by faster reserving cycles. The only way to close it is to measure severity at the demand stage, not the settlement stage. A demand-letter dataset that updates as demands arrive gives the actuary a current severity read that settled-claim data cannot provide. A loss reserve development agent fed with demand-severity trends can project where settled severity is heading, giving the reserving committee a forward-looking adjustment to the historical triangle.

3. How does silent reserve erosion happen when demands outpace reserves?

Silent reserve erosion happens when case reserves are set based on historical settlement patterns while the demands arriving on new claims are systematically higher. The adjuster sets a reserve at the historical average; the demand arrives twenty percent above that average; the reserve is not adjusted to reflect the new severity environment until the claim approaches settlement, by which point dozens of other claims have been similarly under-reserved.

The portfolio-level effect is a reserve deficiency that grows silently for quarters before surfacing in the aggregate numbers. A demand-severity dataset that trends demands against case reserves by line, jurisdiction, and adjuster catches the erosion at the claim level. A loss corridor detection system that compares demand trends to reserve trends flags segments where the gap is widening and the reserves need review.

4. Why are jurisdiction-level inflation differences invisible without demand data?

Jurisdiction-level inflation differences are invisible without demand data because settled-claim severity aggregates across jurisdictions, obscuring the fact that demands in one venue are inflating at three times the rate of another. The portfolio-level severity trend looks moderate while individual jurisdictions are experiencing dramatic inflation that will show up in settlements over the next several years.

This is especially acute in medical malpractice and motor bodily injury lines where venue selection drives outcomes. A demand-letter dataset segmented by jurisdiction reveals which venues are inflating fastest, allowing the reinsurer to load pricing for those jurisdictions specifically rather than applying a uniform inflation assumption across the portfolio.

5. How do obsolete inflation assumptions distort treaty pricing?

Obsolete inflation assumptions distort treaty pricing when the underwriting actuary builds a loss pick using a severity trend derived from settled claims that are two years old, applies it forward, and produces a loss pick that is already inadequate for the demand environment the treaty will actually experience during the coverage period. The treaty is priced to lose from day one.

The pricing gap is largest in lines where demand inflation is most volatile. Directors and officers claims and professional indemnity claims can see demand inflation spike in response to regulatory changes, plaintiff bar focus, or litigation funding trends that are not captured in historical settlement data. A demand-letter dataset that captures those spikes in real time gives the pricing actuary a current view of the severity environment the treaty will operate in.

Stop pricing treaties with yesterday's severity data using Insurnest's demand-letter analytics

Talk to Our Specialists

Visit Insurnest to learn how we help casualty reinsurers extract, structure, and trend demand-letter data for forward-looking severity intelligence.

What do claims analytics leads actually expect from a demand-letter severity dataset?

Claims analytics leads expect a demand-letter severity dataset to extract demanded amounts, injury allegations, jurisdiction, attorney representation, and venue data from every demand letter in the portfolio, update continuously as new demands arrive, segment trends by line and jurisdiction, compare demand trends to case reserves and settlement outcomes, and feed forward-looking severity signals into reserving and pricing workflows.

Thomas leads claims analytics for a casualty reinsurer. His team produces the severity analyses that inform reserving and pricing decisions. The analyses are thorough, well-documented, and backward-looking by design, because the only structured severity data available is the settled-claim data in the reserving triangles. He knows that demand letters contain a forward-looking severity signal that his analyses cannot access because the demand data is unstructured, unextracted, and unanalyzed.

What he wants is a demand-letter severity dataset that sits alongside the settled-claim severity dataset as a second, more current, view of the portfolio's severity trajectory. He wants to be able to tell the reserving committee: "Settled severity shows a six-percent trend, but demand-severity on claims opened in the last twelve months shows an eleven-percent trend. The gap suggests our booked reserves may be light by this amount in these segments."

The expectations are operational and analytical.

• Automated demand-letter extraction at scale. "Pull the demanded amount, injury type, jurisdiction, attorney, and venue from every demand letter without manual data entry." The volume of demand letters in a typical casualty portfolio makes manual extraction infeasible.
• Demand-severity trending by segment. "Show me demand inflation by line of business, by jurisdiction, by injury type, and by attorney representation status." The aggregate trend is interesting; the segment trends are actionable.
• Demand-to-reserve comparison. "For every open claim with a demand, compare the demand to the current case reserve and flag claims where the gap is material." The comparison identifies individual claims that are likely under-reserved and segments where under-reserving is systematic.
• Demand-to-settlement tracking. "For claims that have both a demand and a settlement, track the ratio over time to see whether the negotiation compression rate is changing." If demands are rising faster than settlements, the plaintiff bar is testing higher positions; if the ratio is stable, demand inflation is flowing through to settlements predictably.
• Jurisdiction-level inflation heat mapping. "Show me which jurisdictions are inflating fastest, so pricing can load for those venues specifically." A jurisdiction-level view turns inflation from a portfolio-level assumption into a pricing variable.
• Attorney-firm demand tracking. "Track demand patterns by plaintiff attorney firm, because a small number of firms drive a disproportionate share of demand inflation in many casualty lines." Firm-level tracking reveals concentration risk that portfolio-level analysis obscures.
• Early-warning signals for reserving committees. "When demand trends accelerate in a segment, alert the reserving committee before the next quarterly meeting." The early-warning signal converts the demand dataset from an analytical tool into a governance input.
• Forward-looking severity inputs for pricing. "Feed the demand-severity trend into the treaty pricing model so loss picks reflect the current demand environment, not the settlement environment of two years ago." The pricing function needs current data to price current risk.
• Integration with loss development monitoring. "Connect demand trends to loss development patterns so the actuary can see the full chain from demand to development to settlement." Demand data is the start of the severity chain; development and settlement data complete it.
• Data freshness and update cadence. "The demand dataset must reflect demands received this week, not demands extracted in last quarter's batch run." The value of demand data degrades with latency, just as the value of any early-warning signal does.
• Historical demand data for back-testing. "Build a demand history going back several years so we can test how well demand trends predict settlement trends and calibrate the inflation assumptions we feed into reserving and pricing." The back-test is what gives the reserving committee confidence to act on demand signals.

The real expectation, then, is a demand-letter severity dataset that gives casualty reinsurers a forward-looking view of severity inflation, segmented to the level where reserving and pricing decisions are made, and updated continuously so the data is as current as the claims arriving in the portfolio.

How can reinsurers build a demand-letter severity dataset?

Reinsurers build a demand-letter severity dataset by ingesting demand letters from claim files, extracting structured severity data from each letter using automated document processing, trending demand data by segment and jurisdiction, comparing demand trends to case reserves and settlement outcomes, feeding demand-severity signals into reserving and pricing workflows, and maintaining the dataset as a continuously updated asset rather than a one-time extraction.

Each capability below addresses a stage in turning unstructured demand letters into a structured severity intelligence asset.

1. How does demand-letter ingestion work at portfolio scale?

Demand-letter ingestion works at portfolio scale by connecting to the claims system, identifying claim files that contain demand letters, and pulling those letters into a processing pipeline that extracts the structured data. The ingestion runs continuously, picking up new demands as they are added to claim files.

The technical challenge is document identification. Demand letters arrive in different formats, from different sources, and are stored in different locations within the claim file. The ingestion pipeline must identify which documents are demand letters and which are other correspondence, route the demands for extraction, and log the others. A treaty data quality checker can validate that the ingestion is capturing the expected volume of demand letters relative to the claim population.

2. What does structured severity extraction deliver?

Structured severity extraction delivers a row in the demand dataset for every demand letter, containing the demanded amount, the injury allegations, the jurisdiction, the plaintiff attorney and firm, the venue, the date of demand, and any other fields that the reserving and pricing models consume. The extraction uses a combination of pattern matching, natural-language processing, and document-structure analysis to pull the fields from the unstructured letter.

The extraction accuracy improves over time as the models learn from corrections and as the field catalogue expands. The initial extraction captures the high-value fields, demanded amount and jurisdiction; the matured extraction captures the subtler signals, specific injury language that correlates with ultimate severity, attorney firm identification that enables firm-level tracking, venue descriptions that map to known award patterns. The loss reserve development agent consumes the structured data and integrates it with the claim's reserving history.

3. How does demand-severity trending transform reserving decisions?

Demand-severity trending transforms reserving decisions by giving the reserving committee a current severity signal that is independent of the settled-claim triangle. When the demand trend and the settlement trend diverge, the committee has an early indicator that the booked reserves may need adjustment, and it can investigate the divergence before it becomes a material adverse development item.

The trending runs on the same segment definitions as the reserving analysis: by line of business, by accident year, by jurisdiction. The demand trend for the 2024 general liability accident year can be compared to the settlement trend for the 2022 accident year, and the actuary can assess whether the higher demand trend is likely to flow through to higher settlements or represents a change in negotiation dynamics.

4. Why does demand-to-reserve comparison matter at the segment level?

Demand-to-reserve comparison matters at the segment level because it reveals whether reserves are keeping pace with the demands being made on open claims. A segment where the average demand is systematically above the average case reserve is a segment where reserves are likely inadequate. The comparison quantifies the gap and prioritizes reserve reviews.

The comparison also reveals adjuster-level patterns. If certain adjusters consistently set reserves that are below the demands their claims receive, the pattern may indicate a training need, a system configuration issue, or a portfolio mix difference. The loss corridor detection capability that identifies systematic reserving patterns can be calibrated to the demand-to-reserve comparison, surfacing the adjusters and segments where the gap is widest.

5. How does jurisdiction-level inflation tracking improve pricing precision?

Jurisdiction-level inflation tracking improves pricing precision by replacing a single portfolio-level inflation assumption with jurisdiction-specific inflation rates that reflect the actual demand environment in each venue. The pricing actuary loads a higher inflation rate for the venues where demands are inflating fastest and a lower rate for the venues where demands are stable.

This is the commercial payoff of the demand dataset. Rather than pricing all motor liability treaties with the same severity trend, the actuary can differentiate by jurisdiction, by line, and by attachment point. The differentiation produces loss picks that are more accurate, which means treaties that perform closer to expectations, which means capital that is allocated more efficiently. A historical treaty performance analyzer can measure whether jurisdiction-specific pricing produces better treaty performance than uniform pricing, closing the feedback loop from demand data to pricing accuracy.

6. How does the dataset stay current as demands arrive?

The dataset stays current by running the ingestion and extraction pipeline continuously. As new demand letters are added to claim files, they are picked up, extracted, and added to the dataset within days. The trends, comparisons, and alerts update automatically, so the reserving committee and pricing desk always see the most current demand picture.

Continuous refresh is what distinguishes a demand-severity dataset from a demand-severity study. The study is a one-time snapshot that goes stale; the dataset is a living asset that improves with every new demand. For casualty reinsurers operating in a market shaped by emerging risk forces, where demand inflation can accelerate rapidly in response to legal, regulatory, or social changes, the difference between a study and a dataset is the difference between learning about inflation after it has happened and seeing it as it happens.

Build a living demand-letter severity dataset with Insurnest's document-processing technology

Talk to Our Specialists

Visit Insurnest to learn how we help casualty reinsurers extract, structure, trend, and act on demand-letter data for reserving and pricing.

What does a demand-informed reserving and pricing process look like?

A demand-informed reserving and pricing process runs on two severity datasets instead of one: the traditional settled-claim dataset that anchors the backward-looking view, and the demand-letter dataset that provides the forward-looking signal. The actuary sees both, compares them, and adjusts reserving and pricing recommendations based on the divergence between what demands say is happening now and what settlements say happened two years ago.

Return to Thomas and his analytics team. With the demand-letter dataset in place, his quarterly reserving analysis changes. He still builds the triangles from settled-claim data, because that remains the formal basis for reserving. But alongside the triangles, he presents the demand-severity trends, segmented to match the reserving segments. For segments where demand trends and settlement trends are aligned, the triangles are confirmed. For segments where demand trends are running ahead of settlement trends, he presents an analysis of the gap: how large it is, what is driving it, and what reserve adjustment it implies.

The reserving committee now makes its decisions with two data points instead of one. It can see what the portfolio has already settled for and what the current claims are demanding. The gap between the two, quantified, segmented, and explained, is the committee's best early indicator of where reserves will need to move in future quarters. The committee can act on that indicator now rather than waiting for the settled-claim data to confirm what the demand data already shows.

For the pricing desk, the demand dataset is a direct input. The forward-looking loss pick is built from demand-severity trends in the relevant lines and jurisdictions, adjusted for expected negotiation compression, and validated against historical demand-to-settlement ratios. The treaty pricing model that incorporates demand data produces a loss pick that reflects the severity environment the treaty will actually experience, not the severity environment of two years ago. The treaty is priced to the current market, not the historical one.

Put demand-letter intelligence at the center of your reserving and pricing with Insurnest's technology

Talk to Our Specialists

Visit Insurnest to see how we help casualty reinsurers build demand-letter severity datasets that see inflation before triangles can.

Conclusion

For casualty reinsurers, claim-demand inflation is the severity signal that traditional reserving and pricing frameworks are slowest to capture. Settled-claim data, the foundation of reserving triangles, reflects the demand environment of years past. By the time demand inflation shows up in settlements, it has already been running through the portfolio for multiple reserving cycles, and the booked reserves and priced loss picks are lagging the reality.

A demand-letter severity dataset closes that gap. By extracting demanded amounts, injury allegations, jurisdictions, and attorney data from every demand letter in the portfolio, it creates a forward-looking severity signal that trends can be measured against, reserves can be compared to, and pricing can be built from. The dataset updates continuously as demands arrive, so the reserving committee and the pricing desk always work from the most current severity picture available.

The technology to deliver this exists: automated demand-letter ingestion, structured severity extraction, segmented trending, demand-to-reserve comparison, jurisdiction-level inflation tracking, and continuous refresh. For casualty reinsurers who build it, the demand-letter dataset becomes a competitive advantage in reserving accuracy and pricing precision. For those who do not, demand inflation will continue to arrive as a surprise in the settlement data, quarter after quarter, until the gap between what the portfolio demands and what the models assume becomes too large to ignore.

Frequently asked questions

What is claim-demand inflation in casualty reinsurance?

Claim-demand inflation is the rate at which settlement demands made by plaintiffs are increasing, often faster than general economic inflation and faster than the severity trends embedded in traditional reserving and pricing models.

Why do demand letters need their own severity dataset?

Demand letters capture the plaintiff's opening position before negotiation, settlement dynamics, and reserving judgments compress it. Tracking demands independently reveals the inflation signal that settled-claim data obscures, giving reinsurers an earlier severity read.

How does demand-letter extraction work at scale?

It ingests demand letters from claim files, extracts the demanded amount, injury allegations, jurisdiction, attorney, and venue using automated document processing, and structures the data into a queryable severity dataset updated as new demands arrive.

What signals in demand letters predict ultimate severity?

The demanded amount, specific injury types like spinal surgery, the attorney's track record, venue historical award patterns, and the gap between the demand and the current case reserve are all predictive.

How does claim-demand inflation differ from social inflation?

Social inflation is the broad term for systemic factors driving claims costs higher: litigiousness, nuclear verdicts, third-party litigation funding. Claim-demand inflation is the measurable output captured at the point where plaintiffs price their claims.

Can demand-letter data improve loss reserving accuracy?

Yes, because demand data provides a forward-looking severity signal. When average demands rise faster than case reserves, it signals that booked reserves may be inadequate, giving actuaries an early indicator before the next reserving cycle.

How frequently should demand-letter data be updated for reserving use?

The dataset should update continuously as demands arrive, with trend reports monthly or quarterly. The more current the demand data, the more useful it is as an early-warning signal for reserving and pricing.

What does a demand-letter severity dataset reveal that settled-claim data does not?

It reveals the initial pricing expectation before negotiation compresses it, the trajectory of demand inflation in specific jurisdictions and injury types, and the gap between what plaintiffs demand and what claims ultimately settle for.

About the author

Hitul Mistry is the Founder of Insurnest, an InsurTech company that engineers end-to-end technology exclusively for the insurance industry serving carriers, TPAs, MGAs, brokers, and reinsurers across India, the UAE, and the US. With more than a decade of insurance domain experience, he has built systems spanning underwriting automation, AI-powered underwriting intelligence, claims management, rating and quoting, broking and agency platforms, and reinsurance automation across Health/GMC, Group Life, Motor, P&C, and Reinsurance. Insurnest doesn't adapt generic software to insurance; it builds from the workflow up.

Connect with Hitul on LinkedIn.