Pricing private distressed debt? (advanced analysis)
One of the questions that I had recently during a case study (senior analyst a PE fund) was what price to pay for unlisted debt secured against a hotel (how many cents in the dollar?). Although the debt here is secured on real estate, the rationale I am looking to understand better could equally be applied to any unsecured debt. I have provided the question and outlined my thoughts below but would appreciate if anyone with a distressed background could provide thoughts, comments or suggestions?
[NB: In Distressed Debt Analysis (Moyer), one has the price of the various tranches of debt and one uses these to calculate a YTM under various cases. The difference in my example below is that the debt is not publicly traded so one does not have the current price or YTM etc]
My valuation of the hotels came out at c. $100m vs $80m senior debt and $10m mezzanine debt, which we assume is secured against the property. This implies a total loan-to-value (LTV) of 90%. According, under my base case, all tranches of debt were capital covered and would recover back par (100 cents in the $). Although I estimates that the debt would get fully repaid (even including c. 7% of transaction cost of enforcing ), there is a risk in the downside case (where markets deteriorate) that the debt does not get fully repaid.
Lets look just at the mezzanine here...
Approach A: Probability weighted scenarios
To get the NPV of the debt one could discount the coupons and the expected repayment at maturity at an assumed discount rate. For example, one could assign a:
- 65% possibility to base/upside where you get over $90m for the property (full recovery);
- a 25% downside where the property sells for $85m (50 cents in the dollar); and
- a 10% chronic market meltdown where the property is worth less than $80m and you recover zero
Evidently the probabilities would be based on the expected volatility of the asset (here real estate). Modelling this out would be relatively straight forward. the main challenge would be assigning probabilities to each scenario.
Approach B: Black Scholes
I view this as fundamentally the same as the above, except one is assuming a normal distribution, assuming standard deviation and using the assumed discount rate.
Appropriate discount
One cannot obtain 90% LTV when financing a property in the current market - typically one cannot get more than an LTV of 60% senior debt (which would price at c. %5 all in) and maybe 10% mezzanine on top of that (which would price at c. 12% all in). This tells to me that the mezz (at 90% LTV as per the above) would be trading at a high YTM of above 15% if listed.
That said, the question here is what price the PE should pay (vs likely trading value)? Given that PE target returns of > 20%, my approach would be to use this as the discount rate (or in this case a hurdle rate).
Whether one actually buys the debt is if you can obtain it for under this price (i.e. the price that provides your required rate of return.
...
My questions are can anyone think of any further methodologies? Any guidance would be appreciated!!
Can you please state or clarify the exact question they asked?
Is the security the real estate, the actual hotel operations ex-property or the whole opco/propco package? Which tranche is secured - the senior only? The structure may have a more of less important impact on the price you could pay for either tranche.
With regards to the current valuation: What is the valuation metric and multiple (or rental income + cap rate) that was used to compute the value of 100? How has the valuation of this / comparable companies developed through the cycle and where in the cycle are we right now? This will give you a rough idea of what your bear / base case should look like. Bear in mind that all this should give you a rough idea of the "market value" were this a public security, not the price you should be willing to buy it at.
Required returns: You are right about the discount / hurdle rate: take a look at whatever cash flows the debt holders are entitled to (principal, interest payments) and calculate the price the investor should be willing to pay in order to achieve whatever return they require.
Other considerations: Potential covenant breaches or liquidity issues can result in a default and accelerated repayment of the debt and / or control of equity. This could also impact the price you are willing to pay as you may realize the value before the stated maturity.
Thank you Mad Monkey!! I think I have it now, but would welcome further input!
To clarify, the questions is "What price should you be willing to PAY for the senior and junior debt?" So they are not looking for the market price the debt may be trading at, but at what price could you buy into the security and still achieve your required return on capital. Hence my thinking to use a hurdle rate of 20% [from your comments above, I think you agree with this approach]
My valuation was driven primarily by the NOI / cap rate (derived from my CF model). This was cross referenced against my DCF valuation and a couple of transaction comps (EV / no of rooms).
In terms of security they were both at the opco level and both were secured against the property. Senior debt was first lien and Subordinated debt was second lien. I would presume that the debt also had a floating charge over all the assets but I never checked the detail (was a case study so was very time pressured). I agree this can affect recoveries and influence valuation ranges etc
Your comments re (a) using the cycle as reference point for valuation downside / upsides, and (b) impact of liquidity events are very insightful and will keep in mind for next time!
I found the following analysis useful as well: http://www.distressed-debt-investing.com/2009/09/simplified-distressed-…
Drawing on this, I guess one could model the price you should pay using a sensitivity table: X input = key valuation driver 1 (e.g. cap rate) Y input = key valuation driver 2 (e.g. occupancy rate) Output = Target price of debt = PV of all cash flows (using a 20% discount rate, or maybe even 25% for a margin of safety) i.e. the price you should pay under different scenarios.
To get your answer one would then assigning a probability to each scenario and multiple these by the PVs (or prices) to get the price that should be willing to pay.
Maybe my further questions would be: 1. Any guidance on assigning different scenarios probabilities? (I would lean towards a normal distribution i.e. weighting central scenarios more heavily than the extreme / tail scenarios. This is evidently the hardest and most subjective part of the analysis) 2. Applicable hurdle rate if not stated? (E.g. Distressed Debt investing suggests 25%) 3. Any other method for calculating scenarios for time pressured test conditions? (vs using sensitivity tables)
There are a couple of people that do distressed RE investing on this board that can hopefully help you out.
In my experience, I've seen more things modeled to account for a "margin of safety" rather than based on probabilities of different outcomes. The reality with distressed (real estate debt, never worked with companies) is that there are dozens of outcomes and applying a bull, neutral, bear probability may not give you the best price. Predicting what cash flows are going to be until maturity can be straightforward (given the note is performing or at least paying) and you just need to margin for safety based on your return rate. You can do this in a variety of ways, whether it be discounting the value of the property at exit, raising your return rate, etc.You can sensitize all of this to get your cash-on-cash during the hold, WDP, MOIC, etc.
It's all about the basis on this. The magic question is at what level do you margin for enough error and still win the bid. Again I've never used the probability method mentioned above but I think that leaving a margin of safety in the property level (dispo cap rate, discounting actual NOI for tenants leaving, etc) or the deal level (getting in at a price to easily cover your cost of capital, raising your return rate, etc.) are better ways to margin for risk in the deal.
I am not a lodging or real estate specialist, but reLA's method makes sense when you are "in control" of the equity / real estate.
With debt, if your predictions are right (either everything works out alright and you get repaid at par eventually or you are able to approximate when they would file + for example what a buyer would pay for the entire asset), your cash flows are kinda predetermined, hence you can apply the hurdle rate to get the price you could pay right now.
Of course, you may not get par, e.g. someone buys the assets during bankruptcy for less or it really isn't worth par, and here you may apply various cuts to the exit / terminal / principal value, which will be some sort of margin of safety. Determining a MoS is fairly tricky since a lot of a value gets comes with the repayment of the principal, which is directly correlated to the value of the asset (otherwise, it wouldn't really be distressed). You may not be able to apply your 5% cap rate once you've filed since none of the traditional real estate guys may want to engage in a 363 purchase, hence you may need a 10% or 15% cap rate.
Hence, you may determine a "more realistic / conservative" value that factors in all sorts of issues that may diminish the value of the debt or go for the "blue sky / normalized / arm's length" value and then apply a 30-40% discount. I prefer the former, it gives your thought process more clarity and allows you to audit what went right / wrong ex-post, whereas you don't really know the drivers behind the discount you used. As reLA mentioned, you need to stay competitive and chances are you aren't the only buyer on the block and can ask for any arbitrary discount.
Probabilities aren't great - it's an easy excuse for not putting much thought behind what will need to happen in order to result in the given scenarios, hard to argue with and various distributions can result in all sort of values that can't be well justified.
In your interview case study, I wouldn't worry too much about Monte Carlo and what not - in DD, a lot of things are non-linear and don't follow smooth curves. Focus on rough / conservative guesses of cash flows and a few rough numbers based on information provided and elaborate on other issues one may investigate. Things that spring to my mind is e.g. is the debt solely secured by the RE, does it have recourse to the hotel business, tenor of the two tranches and available cash (maybe they have 15 in cash and the sub is maturing in a month whereas the senior is due in a few years) etc.
Once you have ticked the boxes, this shows how you can genuinely think like an investor and didn't just learn some advanced finance textbook back to front. In the end of the day, your boss will tell you "there is a situation, what do you think we can pay." While you can get a basic operating model going by reading up on the company / industry, you also need to be acutely aware of other things out there that your boss hasn't told you about but that you can dig out and that will have a significant impact on the price you can pay for either tranche.
Aliquam et sunt quis accusamus quia unde. In est consectetur id illo blanditiis autem. Et facere molestiae animi cumque assumenda neque id. A nostrum ipsa provident ipsa reiciendis consequatur sint. Est perferendis impedit magni sit necessitatibus sit. Facilis quia sit error numquam ab minus aliquam. Aut atque blanditiis debitis.
Molestiae voluptates vel repellendus quia distinctio ipsa. Iste aut tenetur voluptatum voluptates. Non tempora voluptate placeat numquam ad asperiores.
Ipsa ea ea aut odit nihil id modi error. Nesciunt eos qui consequuntur odio eius omnis. Non nisi est reiciendis et tempore facilis omnis. Consectetur iste quam similique et vero. Excepturi dolorem minima quas in non harum.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...