Use case · Private credit & asset-backed lending
How do I calculate the expected IRR on a new debt deal before investment committee?
Committee is Thursday. The deal's expected IRR depends on two curves you don't control: the forward curve that projects the floating-rate coupon — or the base rate your margin sits on — and the swap curve that sets your hedge or funding cost on the day you actually hedge. Nearly every loan in the book is floating plus a margin: the margin is yours to negotiate, the curve underneath isn't. Get either one stale, and the margin in the paper isn't the margin you'll realise.
Last updated: July 2026
Project the floating coupons off the live forward curve for the deal's benchmark, price the hedge on the actual amortising notional profile to get your expected funding cost, and combine the two into the IRR — then re-pull the curves at each committee gate so the paper reflects the market on the day it's read. BlueGamma is an interest rate data and pricing platform used by 80+ financial institutions for swap rates, forward curves and cap pricing across 30+ currencies.
How you're probably doing it now
Most deal teams build the committee IRR on one of four rate inputs — a static base-rate assumption keyed in months ago, a weekly rate sheet emailed by the hedge bank, an Excel file of rates that one person exports from the firm's single terminal and circulates, or the terminal itself, kept for exactly this one job.
Here's the circulated-spreadsheet version, from the founding team of a European real-estate private credit fund that writes three or four ten-page committee papers a week for its US institutional backer:
"One person has [a terminal], sends around like an Excel file of spot rates… every week or so… At the moment it's just an Excel file that gets sent around. People are copying and pasting it into models."
— Founding team at a European real-estate private credit fund, all loans floating plus a margin
"No idea when it gets pasted in, who's updated it… when the actual data comes from."
— Same team, on the provenance of the rates inside their deal models
And here's the weekly-email version, as a treasurer running this exact workflow described it:
"The hedge bank sends me an email with a table of one month BBSW… against tenors all the way from three months out to 10 years. And so from that, I'll blend out my curve."
— One-man treasury team at an Australian asset-finance company, originating fixed-rate leases funded floating
"It's not super timely… it's just a bit clunky because it comes via an email."
— Same treasurer, on the weekly rate sheet
The cost of the stale curve is real, and you discover it after the deal is priced — the gap between the rate you underwrote and the rate you eventually hedged at comes straight out of your margin:
"Sometimes I take some margin compression, sometimes I get some margin windfall… I've definitely taken some significant margin compression in the last couple of months."
— Same treasurer
A single spot rate hard-coded in the model months ago. The market has already voted against it — the forward curve is the market's expectation of your coupon, and a flat assumption prices every deal off a number that no counterparty will transact on.
"They think, OK, there's a fixed rate loan, but we'll just take the seven year swap rate and chuck in ten basis points and then feels about right."
— A hedging advisor, on how credit funds without a swaps desk set fixed-rate pricing
Interpolating by hand between the tenors on a once-a-week table also misses curve shape — inversions and dips beyond 10 years that simple interpolation can't capture, and that change the economics of long-tenor deals.
The firm has the data — locked to a single machine and a single login, with the whole deal team queuing behind it.
"I have to get my colleague to download [the terminal] data to then send that over to me… we don't want to buy multiple… licences."
— Structurer at a London private credit fund: one terminal, three deal people
It has the curves — at ~$2–3k/month for a bundle a deal team barely touches, with download caps that bite exactly when deal flow picks up.
"It's just so expensive and so much broader offering than what I needed… it just feels so desperately wasteful."
— The same treasurer, on the premium terminal
How deal teams describe this
The same problem, from private credit funds, direct lenders and deal teams on both sides of the Atlantic:
"[The terminal] is super, super complex and super expensive… I just want a simple user interface, login, download, because I only need Excel data because I'm gonna copy it into my model."
Investment team at a real-estate private credit fund lending across Germany and the Nordics, underwriting margins over EURIBOR, STIBOR and CIBOR
"I own the terminal. It's a separate computer that I have here in the office… When someone needs a request? I'm not always in the office or I might be home sick… It's definitely a struggle."
Same fund: one terminal, a three-person investment team
"We project out our running IRRs and other performance metrics to see how our deals are doing, especially on things that are PIK… and one of the things we need to improve the accuracy is forward curves."
Private credit manager doing direct origination across hard-asset, corporate and real-estate mezzanine lending
"A lot of our models are built on these base rates, since we make a lot of floating rate investments… my team was looking at an investment in Canada, so I was looking for some CORRA forward curves. I wasn't able to find many."
Direct-lending associate at one of the world's largest credit managers, whose private-side team has no terminal access
"You can never afford to have, like, a quote done on even a week or two weeks old… Because so much of your return is coming from getting that right."
Founder of a renewables investor with a £200m commitment, pricing all-equity deals with no financing sense-check
"If there's an improvement in macroeconomics, we keep passing it on to the [investment committee] thinking, you know, things are all good. Meanwhile there is a downward adjustment to the tariff… and now we've overestimated the return."
Project finance advisor in South Africa, on stale inputs surviving weekly committee updates
From forward curve to committee IRR in six steps
Pull the forward curve for the deal's benchmark
EURIBOR, SOFR, SONIA, BBSW — or the benchmarks the free sites barely track: STIBOR, CIBOR, CORRA, SARON, JIBAR. Whichever index the facility floats on, at the right fixing frequency (1M vs 3M vs 6M matters: the gap between them can be ~10bp on the same tenor). Set the valuation date in the curve config so the curve is stamped as-of a specific day. In screening you want the most frequent, most recent information — refresh as often as the pipeline moves.
Project the floating coupon cash flows
Read the forward rate for each interest period and add the deal margin. In Excel, the add-in does this in place: a =BG formula takes each period's start date, end date and the valuation date and returns the forward rate — no re-keying from a PDF, no hand interpolation between tenors, and no copy-paste from a circulated rates file nobody can vouch for. For PIK positions, the same forwards project the capitalising coupon on the accrued balance before you discount back for the running IRR.
Price the hedge on the actual amortising notional profile
If you're originating amortising assets, a bullet swap rate is the wrong number. Upload the notional profile — the amortisation schedule straight from your model — into the swap pricer and get the mid swap rate for that exact profile. That's your expected hedge / funding cost, priced the way the bank's desk will price it.
Combine into the committee IRR
Expected funding cost (the swap rate on your profile) plus projected floating coupons (forward curve + margin) gives you the realised-margin view — and the expected IRR that goes in the paper is built on the market's numbers, not last quarter's assumption.
Re-pull at each committee gate — and archive the curve
Screening, IC1, final approval: refresh the curve at each decision point so the IRR moves with the market rather than surprising you at close. The cadence most credit funds describe — near-daily in screening, then "we kind of stick with a rate when we're closer to the Investment Committee final decision… and then after that we update it at least quarterly" — only works if the refresh is a 30-second re-pull, not a queue behind the terminal login. Keep the as-of-date curve download with the paper as evidence — when anyone asks later why the deal was priced where it was, you can show exactly what the market said that day.
Automate it when the paper cadence is weekly
Writing three or four committee papers a week? Wire the curve in once: a daily API snapshot (get_forward_curve) or =BG formulas that refresh on open, so every model in the fund reads the same timestamped curve — instead of an Excel file of spot rates someone exports from a shared terminal and emails around, with "no idea when it gets pasted in, who's updated it… when the actual data comes from."
For this workflow, side by side
| Static base-rate assumption | Weekly bank email + hand interpolation | Premium terminal (usually one, shared) | BlueGamma | |
|---|---|---|---|---|
| Refresh cadence | Never — set once, drifts | Weekly, when the email arrives | Live — when the person with the login is in the office | Live — 30-second refresh, re-pull at every gate, every seat |
| Niche benchmarks (STIBOR, CIBOR, CORRA, SARON, JIBAR) | No — whatever was keyed in | Only the tenors your bank chooses to quote | Yes | Yes — 30+ currencies, same workflow as EURIBOR or SOFR |
| Curve shape beyond 10Y | No — flat by construction | Missed — simple interpolation can't capture inversions and dips | Yes | Yes — full bootstrapped curve, all tenors |
| Hedge cost on an amortising profile | No | No — call the bank and wait | Yes, if you have the swap-pricing module | Yes — upload the notional profile, get the mid rate |
| As-of-date curve for the committee record | No | Only if you kept the email | Yes — until a download cap freezes the account mid-deal-flow | Yes — set the valuation date, download, archive |
| Straight into the model | Hard-coded cell | Re-keyed by hand from an email | Exported to Excel, emailed around, copy-pasted — no timestamp, no provenance | Yes — Excel add-in =BG formulas or API, timestamped |
| Price anchor | Free (and it costs you at close) | Free — paid for in goodwill and margin compression | ~$2–3k/month per seat + extract limits (curve downloads can be a separately licensed extra) | Monthly flexibility — not an institutional multi-year subscription |
Teams already running committee numbers this way
An asset-finance treasurer — fixed-rate origination funded floating, so hedge timing is realised margin — trialled BlueGamma to replace the weekly emailed BBSW table he blended his curve from. Before trusting it, he benchmarked BlueGamma's historical data against the time series from his hedge bank's email:
"I had benchmarked… the existing time series I have on that weekly email data set against what your platform was giving me in the historical… it was, you know, right on the money. So no concerns on the actual data itself."
— One-man treasury team at an Australian asset-finance company
"Because we're originating amortising leases… I need to build an amortizing profile… I could see that you could load the notional profile… it looked like it had everything it needed."
— Same treasurer, on pricing the hedge against the real amortisation schedule
"We have financial models built every week for different investments… normally I use the swap rates… as a base rate."
Associate at a European private credit fund, on the base-rate input for new deal models"When we are getting a project in its model ready for an investment committee decision, we can recalculate… it's really annoying, we got to call a bank every time we want to update a swap."
Finance lead at a US renewables developer, on re-pricing at committee gates"[We're] kind of just painting everyone with the same brush at the moment."
Analyst at a fast-growing solar-lending fintech, on pricing every deal off one flat rate assumption instead of the curve"When we underwrite [loans], obviously we will rely very much on different forward curves. So we price [loans], we'll basically put on a margin on Euribor or Stibor or Cibor. And for that, we've used [the terminal], which is quite pricey for that small, small task."
Investment team at a real-estate private credit fund lending across Germany and the Nordics"All of our loans are pretty much floating plus a margin and we're just trying to figure out the general return from that."
Founding team at a European real-estate private credit fund, three to four committee papers a week"The requirement for business is only SOFR, SONIA, EURIBOR rates for the current rate and the forecasted rate for the next four quarters."
Head of R&D at a Middle East venture-debt fund, on the curve inputs behind every deal structureFAQs
Use the forward curve for the deal's benchmark (EURIBOR, SOFR, SONIA, BBSW), not a static spot rate. The forward curve is the market's priced expectation of the floating rate over the life of the deal — a static assumption embeds a view the market has already voted against. Some teams use the swap rate for the deal's tenor as a single-number base rate instead; both come from the same curve, and either beats a hard-coded number.
Pull the forward curve for the deal benchmark, read the forward rate for each interest period (period start date to end date), add the deal margin, and apply it to the outstanding notional for that period. In BlueGamma this is a one-click curve download, or an Excel add-in formula that takes the period dates and a valuation date and returns the forward rate directly in your model. For PIK positions the same forwards project the capitalising coupon on the accrued balance before you discount back for the running IRR — the workflow direct-origination funds describe for tracking running IRRs on their book.
A vanilla bullet swap rate will misprice an amortising deal, because the notional-weighted duration is shorter than the final maturity. In BlueGamma's swap pricer you upload the actual notional profile — the amortisation schedule from your model — and get the mid swap rate for that exact profile. That number is your expected hedge cost, and the gap between it and the bank's eventual quote is the bank's margin.
Yes. Set the valuation date in the curve configuration and the platform returns the forward curve as of that exact date. Re-pull at each committee gate and archive the download alongside the paper — if the deal's pricing is questioned six months later, you can show exactly what the market said on the day the decision was made.
The data comes from a leading regulated interdealer broker, one of only three originators of OTC interest rate data globally — the same source the major terminals use — and ties out within roughly 1 basis point. Customers benchmark it against their own bank's rate sheets before relying on it; one treasury team compared BlueGamma's historicals against their hedge bank's weekly table and found it "right on the money."
No. A terminal gives you the curves, but at roughly $2–3k per month for a bundle where a deal team uses a tiny fraction of the content — and the terminal's workflow wants your work inside it, not the data out into your model. In practice most credit funds keep one shared login on a dedicated machine, so every model refresh queues behind whoever holds it. If forward curves, swap pricing on real notional profiles, and fixings are what you need, a focused platform covers the committee workflow on a monthly plan — several teams found BlueGamma by asking an AI assistant for "the best interest rate data provider that doesn't require an institutional two-year subscription."
This is a common gap: SOFR forwards are easy to find on well-known free websites, but the less-tracked benchmarks — CORRA (CAD), SARON (CHF), STIBOR (SEK), CIBOR (DKK), JIBAR (ZAR) — mostly aren't published anywhere free. One direct-lending team's first search was exactly this: a Canadian deal that needed CORRA forwards they couldn't find. BlueGamma covers them the same way as EURIBOR or SOFR — pick the index, set the valuation date, and download the curve or pull it through the Excel add-in or API, across 30+ currencies.
Usually not, for three reasons deal teams cite. Free forward curves typically stop at 10 years, so long-tenor deals end up "just assuming flat" beyond that; the data updates on a lag with no as-of-date history; and there's rarely a commercial-use licence behind it. Teams also simply distrust it — "is it intentionally just a little bit off?… is it still backable by your side?" is how one credit-fund director put it. A paper the committee and later an auditor will lean on needs licensed, sourced data with a valuation-date stamp.
Term SOFR is separately licensed by its benchmark administrator, so BlueGamma provides compounded-overnight-derived SOFR curves rather than the separately licensed term SOFR fixing itself. For most deal models the two track closely — and if your loan is documented on term SOFR while you hedge with an overnight-SOFR swap, the same basis exists in your actual trade, so the model matches reality. We publish a backtest of how the two have moved over time.
Yes, materially. On the same tenor, the difference between a 1-month and a 6-month index curve can be around 10 basis points — real money on a levered IRR. Pull the forward curve for the exact index and frequency in the facility agreement, not a generic benchmark.
Thursday's paper, priced off today's curve
Pull the forward curve, price the hedge on your real notional profile, and put a defensible IRR in front of committee — in the time it used to take to find the bank's email.
Adjacent situations
More scenario guides from the same desk.
.png){kind=logo}