**The following stock may appear in a future issue of The Avid Hog.**

In the last post, Geoff mentioned the use of common sense to estimate the fair value of a stock. In this post, I’ll write about a specific example of using common sense. The stock I’m talking about is **America’s Car-Mart (CRMT)**.

Car-Mart sells used cars in small towns in South-Central states like Arkansas, Oklahoma, and Missouri. But Car-Mart isn’t exactly a car dealer. Car-Mart lends used cars to poor people with limited credit history.

The average car price Car-Mart sells is about $9,721. Customers pay a 7% down payment on average. The remaining balance is financed at a 15% interest rate. Customers go to Car-Mark’s stores to make weekly or bi-weekly payment for the next 29 months. The average bi-weekly payment is about $175.

The focus of Car-Mart is not selling cars. The focus is collecting payments.

When Car-Mart sells a car, the car price is recognized in revenue. The principal balance also increases by the same amount, and allowance for credit loss increases by an estimated amount.

Whenever customers make payments, the part of payment composed of interest is recognized in revenue. The rest of the payment is subtracted from the principal balance.

When a customer defaults, Car-Mart repossesses the vehicle. The fair value of the vehicle is added to inventories. Car-Mart also subtracts the fair value of the vehicle from the remaining principal of the loan. That would be the amount that Car-Mart writes off.

According to these accounting policies, reported earnings are not a good measure of Car-Mart true earnings. Car sales are not cash revenue. Car-Mart can overstate the price of the car they sell to overstate revenue. Or for the same payment stream, Car-Mart can reduce the interest rate to get a higher car price that is recognized in revenue.

Loan originations are not cash spent either. Car-Mart doesn’t lend money to customers to buy cars. Car-Mart lends cars to customers.

This is where we need common sense to calculate owner earnings. The true cash flow in the business is:

CFFO = Receivable Collections - Cost of Goods Sold - G&A - Tax - Increase in Working Capital.

Excluding sales of repossessed cars at costs, Cost of Goods Sold + G&A is about 74% of Car Sales. I'll consider repossessed inventory a type of receivable collections rather than inventory. Working capital is about zero. So, we only need to calculate Receivable Collections.

As we're interested in normal earnings, we'll calculate Receivable Collections in a no growth situation. We're in Year 3 with revenue R. Year 1 and Year 2 both have revenue of R.

Receivable Collections = Collections of Year 1 loans + Collections of Year 2 loans + Collections of Year 3 loans

Collections of loans in each year = collections of successful loans + collections of payments before default + proceeds from repossession.

Car-Mart is lending at 15% and the term length is about 29 months. A $10,000 loan like that with $0 down payment would require 62 bi-weekly payments of $192 for a total of $11,923. So, a successful loan of $P can result in a total collection of 1.19 * P.

CRMT says they normally repossess about 40% of units sold (and over 40% in 2013). The average age of an account at charge-off is 10.6 months. Accounts are on average 71 days past due at the time of charge-off. So, we can say that CRMT can collect 17 bi-weekly payments before default. For the 40% of loans that end up with charge-offs, CRMT can collect 0.4 * 1.19 * R * 17/62 = **0.1305 * R** before default.

The recovery rate is currently 30%. I looked at the industry benchmark and the recovery rate is about 32-33% of remaining principal. The remaining principal of a $10,000 loans after 17 bi-weekly payments is $7,602.70. So, the recovery amount would be 30% * $7,602.70/$10,000 = 22.8% of the face value. A 40% default rate would mean the recovery amount is **0.0912 * R**. Car-Mart’s "inventory acquired in repossession and payment protection claims"/Average Principal Balance was stable at about 11-12%. So, it's conservative to say the recovery amount is **0.0912 * R**.

Let's assume the amount of loan originations to be the same every two weeks. That means total loans generated in each bi-weekly payment is R/26. Excluding the 40% defaulted loans, 60% of loans originated in the first bi-weekly payment will have 10 payments in year 3 for a total of 0.6 * 1.19 * 10/62 * R/26. Similarly, loans generated in the second bi-weekly payment will have 11 payments in year 3 for a total of 0.6 * 1.19 * 11/62 * R/26, and so on.

Total collections of successful loans in year 3 would be 0.6 * 1.19 * 1567/62 * R/26 = **0.6941 * R**

So, Receivable Collections is 0.6941 * R + 0.1305 * R + 0.0912 * R = 0.916 * R.

However, the down payment is 7% of principal instead of 0%. So, Receivable Collections is actually 0.07 * R + 0.93 * 0.916 * R = 0.92 * R

So, CFFO is 0.92 * R - 0.74 * R = R * 18%. As maintenance CapEx is insignificant, pre-tax FCF margin is about 18%.

This calculation has its limitations. First, there can be more than 60% of loans originated in year 2 and year 3 that are still performing in year 3. That would increase collections of successful loans and reduce pre-default collections or proceeds from repossessions. But the net effect would be positive to cash flow in year 3. The bigger limitation is that the total amount of loans originated in each bi-weekly period varies while I assume they're all equal.

If I assume the repossession rate to be 45%, pre-tax FCF margin would be 15.3%. If I assume the repossession rate to be 40% but there are 65 bi-weekly payments, pre-tax FCF margin would be 17%. So, we can see that Car-Mart would make less money with a longer loan term. And the easiest way to improve profit is to reduce the default rate.

It’s hard to estimate Car-Mart’s earnings power accurately. But these calculations show that pre-tax FCF is at least 15% of car sales.

This finding is consistent with Car-Mart’s growth record. From 1995 to 2002, Car-Mart grew sales by 20% annually. From 1998 to 2012, Car-Mart grew sales by 14% annually without using additional debt or equity. The true cash investments in this business are in Cost of Goods Sold and SG&A. A 15% pre-tax FCF margin results in about 9.75% after-tax FCF margin. As COGS + SG&A is about 74% of car sales, 9.75% after-tax FCF margin can fuel 13.18% (9.75/74) growth in COGS + SG&A. From 2003 to 2013, the annual growth rate of COGS + SG&A was 13.1%.

So, 15% is a good estimate of the pre-tax owner earnings margin.

At this point, what is the fair value of Car-Mart? My favorite yardstick is an enterprise value equal to 10 times pre-tax owner earnings. For Car-Mart, that’s equivalent to a ratio of 1.5 times car sales.

But wait! What if Car-Mart can grow 10% over the next 10 years? If that’s true, Car-Mart is worth at least 12 times pre-tax owner earnings, or 1.8 times car sales.

The reason is simple. If today’s pre-tax owner earnings are $1, they will be 1.1^10 * 1 = $2.59 after 10 years. After 10 years, Car-Mart would be worth $25.90 based on an enterprise value of 10 times pre-tax profits. If we pay $12 for Car-Mart today, we would still get an 8% annual return. Also, at a 10% growth rate, Car-Mart would have excess cash to repurchase shares. That would add 1% or 2% to our investment return.

But again, don’t worry too much about what’s the fair multiple. It can be 12, 13 or 14 times pre-tax owner earnings for Car-Mart. The bigger questions are about qualitative factors. Is Car-Mart durable? Does Car-Mart have a strong competitive position? If yes, is the competitive position sustainable? How likely is it for Car-Mart to lose underwriting discipline? Is Car-Mart’s capital allocation good or bad? And does Car-Mart have favorable future prospects?

If the answers to all these questions are positive, we know that we would get a good bargain paying an enterprise value of 10 times pre-tax profits.

**Talk to Quan about America’s Car-Mart (CRMT)**

** **

*Try Before You Buy**: To sample the current issue of Geoff and Quan’s newsletter, The Avid Hog, just email Subscriber Services*

*and ask for a copy.*