Data is still lacking on how insurers in emerging markets will fully adapt—Colombia, Kenya, even parts of Eastern Europe are still wrestling with implementation. And honestly? Experts disagree on whether the long-term benefits outweigh the short-term chaos. But one thing’s certain: if you're reading financials post-2023, you can't afford to ignore how remaining coverage liabilities warp reported profits, especially in life insurance.
Understanding the Core: What Even Is Liability for Remaining Coverage?
At its most basic, the liability for remaining coverage captures what an insurer owes for services not yet delivered under an insurance contract. Think of it like this: if a policyholder pays $1,200 for a 12-month car insurance plan, after three months, 75% of the coverage remains. But under IFRS 17, it’s not just about time—it’s about risk, capital, and expected claims. This isn’t prorated like GAAP. It’s a living model, recalculated each reporting period.
Sure, you could call it a deferred revenue analog—but that’s oversimplified (and actuaries hate that). The real kicker? It includes a risk adjustment. That means if a storm season worsens, or mortality spikes post-pandemic, that liability can jump overnight. And because the liability must reflect current estimates—not initial assumptions—it introduces volatility no one saw coming in 2010.
How It Fits Within the IFRS 17 Framework
The liability for remaining coverage sits inside the fulfillment cash flows, which also include the residual margin and risk adjustment. It’s distinct from the liability for incurred claims—those are for events already reported. Remaining coverage? Pure future obligation. It grows when new contracts are signed, shrinks as services are delivered. But it doesn't decline linearly. That’s where people get tripped up.
The Role of the Contractual Service Margin (CSM)
The CSM is essentially unearned profit. It starts as zero or positive and gets unlocked over time as services are provided. But—and this is critical—if actual experience differs from assumptions (say, fewer claims than expected), the CSM gets adjusted upward. That adjustment flows directly to profit, even if no cash changed hands. This mechanism is why some insurers saw 30% swings in net income during early adoption (Allianz reported this in Q1 2023). You don’t see that under IFRS 4. And that’s exactly where the market gets spooked.
Why the Liability for Remaining Coverage Changes Everything in Financial Reporting
It used to be that insurance profits were predictable, almost sleepy. Now? One quarter, you’re up; next quarter, down—because discount rates shift or mortality tables get revised. A 50 basis point drop in risk-free rates can inflate liabilities by 5–8% across large portfolios. That’s not theoretical. AXA saw a €1.3 billion impact in 2023 from rate changes alone. The issue remains: investors aren’t trained to parse this kind of volatility. They look at EPS dips and assume trouble. But often, it’s just accounting mechanics.
And yet, the standard insists on this transparency. Because hiding risk in reserves was part of what made Solvency II necessary in the first place. So IFRS 17 flips the script: show the uncertainty, don’t bury it. But because markets hate uncertainty, stock prices wobble. We’ve seen it in Japan, where Tokio Marine’s shares dipped 4% post-earnings despite solid underlying performance. Perception matters as much as math.
Impact on Profit Recognition Timing
Under old rules, insurers could recognize profits early. Now? Profits emerge gradually, as the CSM is amortized. This means a policy sold today might contribute 5% of its margin in year one, 15% in year two—depending on coverage pattern. A level premium term life policy, for instance, might front-load cash flows but back-load profit recognition. That’s counterintuitive for non-actuaries. And that’s where confusion sets in.
Volatility in Equity and Net Income
One insurer reported a 70% drop in quarterly net income—not because claims spiked, but because discount rates fell. The liability for remaining coverage ballooned, CSM adjustments couldn’t compensate, and losses hit the P&L. But equity? It absorbed some via OCI, depending on jurisdiction. In the U.S., for foreign subsidiaries, the noise is amplified because of translation effects. The problem is, most analysts still treat insurance like a stable yield play. They don’t realize the accounting is now more like a derivative model.
Calculating the Liability: Assumptions, Models, and Real-World Noise
You need four things: expected future cash flows, discount rates, risk adjustment, and coverage units. Sounds clean. But each layer is messy. Expected premiums? Easy. Expected claims? Not so much. For a 20-year whole life policy, you’re projecting mortality, lapses, expenses, inflation—all over decades. And each assumption is interdependent. Change lapse rates, and your cost allocation shifts. Change discount rates, and the whole liability surface warps.
Actuarial teams now run thousands of simulations weekly. Swiss Re uses AI to stress-test 12,000 parameter combinations per contract group. And that’s before considering currency risk—say, if you’re pricing in euros but paying claims in Kenyan shillings. Which explains why some CFOs are quietly lobbying for simplified approaches, especially for short-duration contracts. But the standard allows that only in narrow cases. In short, complexity is baked in.
Discount Rates: The Silent Driver of Volatility
They’re based on high-quality corporate bonds—rated AA or better. But here’s the catch: you can’t hedge perfectly. Bond markets jump. Central banks pivot. And when the ECB cut rates by 25 bps in September 2023, several European insurers had to revalue liabilities within 72 hours. Because the present value of future claims increased. That said, some firms use “laddered” discount curves to smooth the effect. It’s allowed, but it’s controversial—some auditors push back hard.
Risk Adjustment: How Much Buffer Do You Really Need?
This isn’t a fixed percentage. It’s the compensation an insurer would demand to take on the uncertainty of future cash flows. Methods vary: cost of capital, confidence level, even market-perceived risk. A health insurer in Brazil might set it at 6%, while a German auto insurer uses 3.5%. Why? Different regulatory cultures, different litigation risks. But because it’s judgment-heavy, it’s a red flag for auditors. KPMG flagged 22 discrepancies in 2022 filings across Latin America alone.
IFRS 17 vs. IFRS 4: The Before and After Shock
Under IFRS 4, insurers could use “building block” models with smoothing. Profits were stable. Reserves were padded. And no one really knew the true risk exposure. Fast-forward to IFRS 17: every assumption is scrutinized, every fluctuation reported. It’s like switching from a sketch to an MRI scan. You see more—but some of what you see is unsettling.
Take Aviva. In 2017, under IFRS 4, their life division showed steady 8% ROE. By 2023, under IFRS 17, ROE swung between 5% and 11% quarterly. Same business. New accounting. Because now, when longevity improves (good for solvency), it hurts short-term profits—liabilities increase. The irony? Good news looks like bad news. That’s the paradox no one saw coming.
Profit Recognition: Deferred and Distorted
IFRS 4 let you recognize margins early. IFRS 17 forces deferral. So a product that was “profitable” in year one under old rules might show zero profit until year three. That skews internal KPIs, frustrates sales teams, and confuses investors. Some firms, like Ping An, now publish dual reporting—both standards side by side. But that’s not sustainable. Eventually, markets must adapt.
Disclosure Depth: From Opaque to Overexposed?
Now you must disclose CSM roll-forwards, risk adjustments, discount rate sensitivity, and coverage unit methods. A typical annual report section runs 40+ pages. That’s transparency, sure—but is it useful? A fund manager I spoke with put it bluntly: “I just want to know if the company is making money. I don’t need the actuarial autobiography.” There’s a real risk of information overload.
Frequently Asked Questions
How does the liability for remaining coverage affect quarterly earnings?
It can inflate or deflate profits based on changes in assumptions. If mortality improves, life insurers must increase liabilities—hurting earnings. If discount rates rise, liabilities shrink, boosting profits. But this is non-cash. And because the CSM adjusts, the effect isn’t always immediate. Still, the swings are real enough to move markets.
Can the liability go negative?
No—but the CSM can. If future losses are expected, the CSM absorbs them first. If it hits zero, losses hit the P&L immediately. That’s called “onerous contracts.” During the pandemic, several travel insurers hit this wall. Their remaining coverage liability didn’t turn negative, but their CSM did—and losses poured through.
Is this liability the same across all insurance types?
Not even close. For long-duration life policies, it’s complex, multi-decade modeling. For motor insurance? Often simplified using premium allocation approach. Short-duration contracts (under one year) can use a different method. But even then, risk adjustment and discounting apply. So no free passes.
The Bottom Line
The liability for remaining coverage isn’t just an accounting line—it’s a philosophical shift. It says: stop hiding behind smoothing, stop deferring reality. Face the risk. I find this overrated in calm markets but invaluable during crises. Because when the next pandemic hits, we’ll know—fast—where the real exposure lies. That said, the standard could use more flexibility for small insurers. Automated modeling at this level costs $2–5 million to implement. Not every company can afford it. So while the intent is sound, the execution risks leaving smaller players behind. And that’s a problem no model can fix.