From Look-Through to Look-Across: The Next Stage of Insurance Investment Oversight

The short answer: look-through examines each holding in isolation — what does this instrument actually own? Look-across examines the relationships among holdings, counterparties, and collateral: duplicated exposures held through different legal forms, affiliated dependencies, correlated collateral, and concentrations that exist only in aggregate. The practical tool is an enterprise exposure map keyed by ultimate economic parent rather than legal issuer. This is Part IV of the Capital Credibility Series, closing the arc begun in Part I, Part II and Part III.

What look-through cannot see

The NAIC's 2026 capital reforms — differentiated treatment for collateral loans, new C-1 factors for CLOs — completed the industry's long argument about look-through: the principle that an instrument should be assessed on what it actually owns rather than what it is called. But adopting look-through has exposed the question it cannot answer. Consider a portfolio that has done everything right by current standards. Every CLO tranche is analysed to its loan pool. Every collateral loan is classified by what secures it. Every fund interest is decomposed. Each vertical analysis is sound — and the portfolio can still hold the same ultimate borrower through three different wrappers, the same sponsor's origination platform behind four different asset classes, and collateral pools across two reinsurance trusts that would reprice together because they share tenants, managers, or financing channels. Issuer limits do not catch this, because the issuers differ. Asset-class limits do not catch it, because the classes differ. The exposure is real, material, and formally invisible: it exists only in the relationships.

What look-across asks

Look-across poses four questions of an enterprise rather than an instrument. Where are we exposed to the same economics more than once — duplicated exposures held through different legal forms? Where do our assets depend on affiliated parties — managers, originators, or sponsors connected to us or to each other? Where would supposedly independent collateral move together — correlated collateral behind different counterparties? And what concentrations emerge only in aggregate — hidden accumulations that no schedule, taken alone, reports? These are the same instincts Part II applied to circular structures and Part III applied to sponsor-dependent ones, widened to the whole balance sheet: the unit of analysis stops being the instrument and becomes the economic relationship.

Building the map

The practical version of this is an enterprise exposure map, and the discipline is less exotic than the name suggests. It joins three inventories most organisations already maintain but never connect: the statutory investment schedules, the reinsurance counterparty list, and the collateral and trust asset registers. Each entry is then keyed not by legal issuer but by ultimate economic parent — who originates, who manages, who owes, who sponsors — the same five-question trace this series proposed for circularity, applied systematically. The output is unglamorous: a table showing, for each economic group, every place it touches the enterprise — as investment, as counterparty, as collateral, as service provider. The first draft is always wrong and always revealing. Groups nobody thought of as concentrations appear in a dozen rows; exposures the investment guidelines carefully capped resurface uncapped inside collateral schedules.

Who owns it

Look-across fails as a project and works as a routine, and it needs a single owner because its raw material crosses departments. Investment teams hold the schedules, reinsurance teams hold the counterparty relationships, legal holds the trust agreements — and in most organisations no one holds the joins. Whether the owner is the chief risk officer, the investment committee, or (in a captive) the board itself matters less than the assignment being explicit, the map being refreshed on a calendar, and its findings having somewhere to go: into investment guidelines as relationship-based limits, into reinsurance due diligence as counterparty questions, into collateral review as substitution triggers.

Why this will not stay optional

Regulators built look-through into the capital framework when it became clear that legal form was concealing economic substance at the instrument level. The current examination of interconnected and affiliated structures is the same logic arriving at the portfolio level, and it is reasonable to expect supervisory questions to follow the same path — from "what does this instrument hold" to "what do your holdings have in common." Organisations that build the map now will answer those questions from their own analysis. The alternative, as this series has argued from the start, is being explained to — by a regulator, a rating agency, or a counterparty — about your own balance sheet.

This concludes the current arc of the Capital Credibility Series: capital rules that price economics rather than labels, a test for structures that reference themselves, a standard for structures no one can independently explain, and a map for the exposures that only exist in aggregate. The common thread is one sentence: credibility is the ability to demonstrate, in your own words and your own analysis, that the capital is what it claims to be.

Educational commentary on insurance capital and regulatory developments, current as at August 2026 and subject to change. Not insurance, actuarial, tax, legal, or financial advice for any specific organisation.