Four haircuts, each simple, whose product takes a guarantee to about a third of its face.
A personal or sponsor guarantee is worth the claim it supports, times the probability the guarantor can still pay, times the share a settlement or judgment actually collects, less the legal cost, discounted over the years it takes to collect. On a €30.0M defaulted loan with a guarantee capped at €6.0M, that is €2.17M today, 7.2 points of recovery against the 20.0 points the cap suggests. The guarantee is real money, but about a third of its face.
Worked in full in The Real Estate Workout by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
Workout teams routinely add the guarantee cap to the enforcement recovery and call the total "our position". That treats a contested, delayed, unsecured claim on a person or a holding company as if it were cash. The four haircuts below are each simple; it is their product that surprises.
| Input | Value |
|---|---|
| Loan outstanding | 30.00 |
| Expected shortfall after enforcing on the property | 9.00 |
| Guarantee cap | 6.00 |
| Probability the guarantor is still solvent when called | 70.0% |
| Share of the claim collected, if solvent | 75.0% |
| Legal and enforcement costs against the guarantor | 0.40 |
| Years from call to cash | 2.5 |
| Lender's discount rate | 10.0% |
The discount rate is higher than one would use for the property, because a claim against an individual or a holding company carries more uncertainty than a building. The 75.0 per cent collection share stands for the settlement a guarantor will offer rather than defend: guarantors negotiate, and lenders usually accept a discount for speed and certainty.
Guarantee value = [min(cap, shortfall) × P(solvent) × collection share − legal costs] × (1 + r)−t
In Excel: =MAX(MIN(Cap,Shortfall)*PSolvent*Collect-Legal,0)/(1+r)^t. The MIN matters: a guarantor owes the lesser of the cap and the actual loss, so a smaller shortfall shrinks the claim.
| Step | Lost |
|---|---|
| Guarantor insolvent when called | 1.80 |
| Settlement discount | 1.05 |
| Legal costs | 0.40 |
| Time value over 2.5 years | 0.58 |
| Value to the lender today | 2.17 |
| Cap | 6.00 |
The guarantee adds 7.2 points of recovery on the 30.00 loan, not 20.0. It is worth 36.1 per cent of its cap. The largest single loss is solvency: a guarantee from a sponsor with the rest of its wealth in property is a correlated asset, weakest in the scenario in which it is needed.
The guarantee's value is also its release price. In a discounted payoff the borrower almost always asks to be released from the guarantee. The 2.17 is the least the lender should charge for that release, on top of the payoff for the loan itself. A lender that accepts a release fee of 1.50 has given away 0.67 of value without noticing.
| Change | Value | Points of the loan |
|---|---|---|
| Solvent with probability 90.0% | 2.88 | 9.6 |
| Solvent with probability 70.0% | 2.17 | 7.2 |
| Solvent with probability 50.0% | 1.46 | 4.9 |
| Solvent with probability 30.0% | 0.75 | 2.5 |
| Collected in 1.0 year | 2.50 | 8.3 |
| Collected in 4.0 years | 1.88 | 6.3 |
| Shortfall only 4.00 (below the cap) | 1.34 | 4.5 |
Solvency moves the answer most: each 20 points of probability is worth about 0.71, or 2.4 points of recovery. Time matters less than people fear, because the legal process is slow but a settlement usually is not. And a guarantee is worth less when the property recovers well, since a 4.00 shortfall caps the claim at 4.00. The legal costs are a fixed toll: below an 8.9 per cent probability of solvency they consume the whole expected collection, and pursuing the guarantor destroys value.
The common mistake is to add the cap to the property recovery. A credit paper that says "enforcement recovers 70 per cent, plus a 20 per cent guarantee" has promised 90 cents where the arithmetic supports about 77. The second mistake is to treat the guarantee as independent of the property: the probability of solvency should be taken in the downside scenario, not from the guarantor's statement of net worth at origination. The third is to read a non-recourse carve-out guarantee as a payment guarantee. A carve-out only springs on the events listed in the agreement, and if none occurred, its value is zero.
The guarantee sits alongside the enforcement value when pricing a discounted payoff, and alongside the success and failure branches when deciding whether to extend or enforce.
Value a guarantee as a probability-weighted, discounted, net claim: 6.00 of cap becomes 2.17 today, 36.1 per cent of the face. Put that number, not the cap, into the recovery, and charge at least that much to release it. The free workbook for this case prices the sponsor guarantee alongside the extension, with the solvency discount, the time discount and the years to collect as inputs.
No. The cap is a ceiling on a contested, unsecured claim. In the worked example a 6.00M cap on a 30.00M loan is worth 2.17M after insolvency risk, a settlement discount, legal costs and 2.5 years of waiting, so a paper claiming 70 plus 20 points of recovery should say about 77.
At least the guarantee's present value to the lender, on top of the payoff for the loan. In the illustrative case that is 2.17M. A lender accepting a 1.50M release fee gives away 0.67M, because it values the release against nothing rather than against the guarantee it surrenders.
No. A carve-out guarantee only springs on the events the loan agreement lists, such as fraud, unauthorised transfer or voluntary insolvency. If none occurred it is worth zero, whatever the size of the shortfall. A capped payment guarantee responds to the loss itself, here worth 7.2 points on the loan.
This article is one calculation from The Real Estate Workout. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: real estate investing, finance and fund management → · All 681 articles →
If this book helped, or didn’t, a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.