Tinlake Pricing and Valuation Series — Part 2: Valuing an asset portfolio

In the first part of our Tinlake pricing series, we have taken you on a tour with our partner ConsolFreight to price real-world assets. If you haven’t read it, we recommend doing so. Based on a scorecard combining different pricing factors, we have determined finance fees and advance rates for invoice backed assets financed through Tinlake. Tinlake is our open, smart-contract based marketplace of asset pools bringing together asset originators, such as ConsolFreight, and investors that seek to utilize the full potential of Decentralized Finance. We will now take the next step in our pricing journey: Assuming these invoices would have been financed through Tinlake, how could these financings be valued?

Valuation is the process of determining the current worth of an asset by assigning a monetary value. The value of a portfolio of assets is often also expressed as the net asset value (NAV). When would you need to determine the value of an asset portfolio or the NAV? Foremost when this portfolio is sold or when investors want to join/exit an existing investment pool. Then the portfolio value ultimately determines the investment/redemption price. Note that for these purposes the portfolio value may be different from the book value or accounting value of a portfolio.

It’s all about expected cash flows

Determining the value of illiquid assets is notoriously difficult because — by definition — there isn’t a liquid secondary market to determine the value, unlike many stocks, bonds, or most fungible tokens. For illiquid asset portfolios, the valuation methodology is thus often based on a fair value valuation utilizing a financial model (“marked to model”). This often comes down to valuing the present value of future cash flows expected to receive based on these financings — the so-called discounted cash flow (“DCF”) method.

Revisiting ConsolFreight

To understand how this works we will have a look again at ConsolFreight. ConsolFreight is a SaaS freight forwarding technology provider that advances working capital finance to freight forwarders’ invoices and collects from shippers. Let’s assume ConsolFreight has just financed three invoices with different freight forwarders with an invoice amount of USD 1,000 and due dates in 30 / 60 / 90 days. The advance and finance fees to finance these assets in Tinlake are taken from the scorecard we derived in the first blog post:

We further assume that we have one invoice from each of the risk groups A-C translating into advance rates of 95% / 95% / 90% and finance fees of 10% / 12% / 14%. The advance rate determines how much of the invoice amount of 1,000 can be financed. To advance USD 950 / 950 / 900 to each freight forwarder, ConsolFreight tokenizes their invoices and uses the non-fungible tokens (NFTs) as collateral to finance these invoices through Centrifuge Tinlake in DAI. If you wanted to buy this portfolio today, what would this portfolio be worth to you? What would you be willing to pay today to receive the cash flows expected in the next 30–90 days from these assets? Or in other words, what is the present value that you assign to this portfolio?

Building a simple valuation model

To illustrate how such a portfolio could be valued, we will develop a very simple DCF model based on the cash flows we expect to receive when the assets financings are repaid. Tokenized invoices are good examples as they have relatively simple cash flow structures — one advancement and one bullet repayment. Valuation models for other types of financing structures with several fee payments and repayments would obviously be more complex but still could be valued with the same approach.

Step 1 — Derive the expected cash flows

To derive the cash flows originating from these assets, the natural starting point is to look at the contractual obligations associated with the financing. These will often tell you when to expect which payment or cash flow. As mentioned, invoice financing comes with a very simple cash flow structure and with the bullet repayment usually coinciding with the due date of the underlying invoice. Thus we can expect the repayment of the three financings in 30, 60 and 90 days. As most DeFi protocols do, Tinlake compounds interest every second. Compounding the three financings of DAI 950 / 950 / 900 with a 10% / 12% / 14% annual interest rate every second gives us expected repayment amounts or cash flows of roughly DAI 960 / 966 / 932 after 30 / 60 / 90 days:

Step 2 — Adjust for risk

The next step is to adjust these expected cash flows by different kinds of risk depending on the asset type. In practice, these may be e.g. credit risk, risk of early payment, or fraud risk. Usually operational costs to manage the portfolio and write-offs for late payments would be considered as well. For simplicity, we will only focus on credit risk here and use a common “expected loss” approach to model it. The expected loss of a future cash flow is calculated by multiplying the probability of default (“PD”), so the probability of not getting paid back in full, with the loss given default (“LGD”) which defines how much of the expected payment is actually lost in case of a default. The background of including the LGD is that often payment obligations are at least partly met or a collection agency is able to recover part of the outstanding amount. Note that the LGD is diametrical to the so-called recovery rate (Recovery rate = 1 — LGD).

PDs and LGDs vary heavily between different asset classes and industries. The factors driving PDs and LGDs would be similar to the factors we introduced in the first blog post. For invoices, PDs and LGDS are historically relatively low with e.g. PDs ranging between 1–2%. This is because many of the invoice repayments can be at least partially collected. PDs are usually positively related to finance fees, which means that higher credit risk in terms of higher PDs carries higher finance fees to compensate for this additional risk. PDs and LGDs are often calculated based on historical defaults and calibrated per rating class.

To continue with our example from above, we’ll add the following PDs and LGDs to our scorecard:

Multiplying the expected repayment amounts of DAI 960 / 966 / 932 for each financing with corresponding PDs and LGDs gives us expected losses of DAI 3.8 / 1.2 / 4.7. To adjust the expected cash flows by credit risk we simply subtract the expected losses from the expected repayment amounts:

This risk-adjustment may seem relatively small on an asset level, but remember that this approach is supposed to work on a portfolio level. If you have a portfolio of 1,000 financings these relatively small amounts of individually expected losses will add up and normally be relatively close to the actual loss for the entire portfolio.

Step 3 — Discount the risk-adjusted expected cash flows

The final step is to discount the risk and cost adjusted expected cash flows with an appropriate benchmark rate to derive the present value (“PV”). Why? Because from a financial perspective, time really is money. The time value of money is a basic financial concept that holds that money in the present is worth more than the same amount of money to be received in the future. This holds because the money that you own right now can be invested and earn a return, thus creating a larger amount of money in the future. To account for this effect we need to discount the expected cash flows with an appropriate benchmark rate. This is usually either a risk-free rate or a rate with a similar level of risk. For simplicity, we chose a discount rate of 10% here. Note that we make no further adjustments to the discount rate as we explicitly considered credit risk before.

The standard formula to calculate the present value for an expected future cash flow is

with r being the discounting rate and t the period of the cash flow. As we deal with intra-year cash flows, the formula becomes

with n being the number of discounting periods per year, e.g. 360 days for a financial year. If we now discount the three risk-adjusted cash flows of DAI 956 / 965 / 927 expected to materialize in 30 / 60 / 90 days with a discount rate of 10%, we get present values of DAI 948 / 949 / 905 for the three financings.

To derive the present value of the financed portfolio we simply add the three individual PVs and get a fair value for our portfolio of DAI 2,802. The following table summarizes the entire approach:

Outlook

We are not finished yet. The most important questions still remain unanswered. How would this Portfolio NAV be translated into prices for Tinlake’s two tokens, DROP and TIN? And how can the prices and valuations be brought on-chain and into Tinlake? These questions will be answered in the next parts of Tinlake’s Pricing and Valuation Series, so stay tuned.