- Home
- Case studies
- Infrastructure investor
An infrastructure investor reallocating when the market moves.
An infrastructure investor, invented for this page, runs a real-assets programme across fibre, water, renewables, storage, toll roads, an airport stake and data centres. When a market event lands, the questions are always the same: which positions are exposed, which limits are close, and what should we do about it, approved by whom?
The situation
A programme that looks diversified on a sector chart
Every figure on this page is invented. The programme holds 11 positions across three vintages: fibre networks, a regulated water business, onshore wind and solar, battery storage, two toll roads, a minority airport stake and two data-centre campuses. Net asset value is $6.4bn on $7.2bn committed, with $1.3bn still unfunded and three deals in the pipeline.
Its exposures are shared in ways a position-by-position review misses. Interest rates, power prices, inflation and credit spreads move several positions at once, and the United Kingdom is more than a third of the book. Three of the assets are still under construction.
What Capibud does
From a headline to an approved reallocation
- Read the documents as they arrive. An un-templated asset model, an investment committee memo, a quarterly review deck and a scanned valuation report. Where an independent valuation and a deal-team mark disagree, both are shown and the independent valuation takes precedence; they are never averaged.
- Put every position on shared factors. 2,500 scenarios across eight market factors: rates, credit, power prices, inflation, equities, build costs, traffic and policy.
- Replay market events. Each event, such as a central-bank rate surprise, a power-auction result or a change to a renewable support scheme, is stored with its date, its source and its measured market move. It marks the exposed positions and re-runs the scenarios.
- Check limits and stresses. Sector, region and single-name limits, historical stresses, reverse stress, liquidity coverage and the denominator effect on the parent's allocation band.
- Propose and route. Breaches lead to a proposed reallocation (sell-downs, hedges or slower commitment pacing), routed for second-line concurrence, then to the head of infrastructure below $150M, the CIO below $600M and the investment committee above. The approved change is applied and recorded.
- Look through to construction. Assets under construction are tracked with Execute, so a slipping campus or road widening shows up in the allocation view.
A scheduled feed reads public filings and price series. When a series moves past a threshold the investor has set, Capibud flags the exposed positions and re-runs the scenarios.
Results
Close to the edge, but not over it
Illustrative example — figures invented for illustration; not a client or a Capibud demo. Utilisation is the position's share divided by its limit.
The whole programme on one set of scenarios
| Measure | Value | Reading |
|---|---|---|
| Expected gain | $0.52bn | About 8.1% of NAV |
| Bad year (P10: nine scenarios in ten do better) | −$0.47bn | About 7.3% of NAV |
| Illustrative rate-and-inflation shock | −$0.93bn | About 14.5% of NAV; largest hits in toll roads, then renewables |
| Reverse stress: move that exhausts appetite | ≈ 2.9σ | Roughly 1 in 150 years; led by rates and power prices |
| Liquidity coverage, next 12 months | 1.45× | Against a 1.25× minimum |
| Private-markets weight after a 25% equity fall | 13.8% | Peaks in 2029, inside the 10–15% band |
Illustrative example — figures invented for illustration; not a client or a Capibud demo. Return periods would be model-based.
Construction assets, looked through
| Asset | P80 overrun | P80 slip | Verdict |
|---|---|---|---|
| Toll-road widening | +16% | about 8 months | Escalate |
| Data-centre campus · power and cooling | +11% | about 4 months | Recover |
| Battery storage site · grid connection | +7% | about 9 weeks | Watch |
| Fibre build-out · phase 2 | +3% | days | Proceed |
Illustrative example — figures invented for illustration; not a client or a Capibud demo. P80 is the figure the overrun or slip stays within in four scenarios out of five.
Limits
What this example can and cannot tell you
- The figures are invented. Every number on this page was made up to show how the method works. It is not a client, not a Capibud demo workspace and not drawn from any investor's data.
- Event effects are estimates. An event's market move can be measured from public data, but how it moves a programme's factors is a labelled estimate.
- Market data needs a licence. Public sources such as FRED and SEC EDGAR are fine for demonstration. Licensed market data is required before any customer use.
- No trades. A reallocation changes a plan inside Capibud. Capibud does not move funds or place orders.
What this means for you
Concentration is a question about factors
One book, shared factors
Fibre, water and renewables can look diversified on a sector chart and still move together on rates, inflation and power prices.
Events with provenance
Every proposal starts from a dated event with a source, and every number it shows comes from a recorded run.
Authority built in
Reallocations route by size through the approvers you already have, with second-line concurrence first.
Related
Use cases and industries in this example
Allocation uses Invest; the look-through uses Execute. The packs below hold infrastructure baselines.
Invest: venture and fund portfolios
Tickets, reserves and concentration limits on correlated scenarios.
Read the guideA climate fund: building around the power law
Three ways to deploy $300M, compared on identical scenarios.
Read the caseA data-centre developer: phase 1 in construction
What a look-through asset looks like from the developer's side.
Read the caseBring us up to 15 holdings.
An Investor Sprint is a four-week diagnostic of concentration, correlation and capex exposure on your own book.