Description

Business Model

 Both JDG and LEW sell furniture, appliances and electronic equipment to the middle and lower income markets in South Africa. These items are sold on credit and cash.

The key to the trade, however, is not to view these companies as retailers, but, as micro lenders who are effectively swapping furniture for a loans.

The math of a typical credit transaction would be to swap furniture advertised at $10 000 for a 24 month instalment loan of $1 000 per month. Using a 12% discount rate this is equivalent to swapping $10 000 for $21 000. Knock-off a through-the-cycle average of 10% for bad debts and the “retailer” is swapping $10 000 for $19 000. This 1.9 conversion ratio is key to unlocking the arbitrage opportunity described below.

LEW is expected to earn R7.00 over the next 12months. At the R48 share price LEW is trading at a rolling forward 12month PE = 6.8.

JDG is expected to earn R4.28 over the next 12months. At the R36.65 share price JDG is trading at a rolling forward 12month PE = 8.6.

Because of the distortions created by GAAP it is better to build a model from the bottom-up to determine the NPV that is created from 1 year of trading.

NPV created by LEW

The following is extracted from the 12month income statement ended September 2008.

Credit Sales              R1 239mil

Cash Sales                R   644mil

Cost of Sales            R1 277mil

Oper Expenses         R1 294mil

Using the 1.9 factor described above the

NPV created = (R1 239*1.9) + R644 - R1 277 - R1 294

NPV created by JDG

The following is extracted from the 12month income statement ended August 2008.

Credit Sales              R3 061mil

Cash Sales                R6 214mil

Cost of Sales            R6 627mil

Oper Expenses         R4 292mil

The NPV created = (R3 061*1.9) + R6 214 – R6 627 – R4 292

EV Multiple

LEW’s market cap = R4 250mil, net debt = R765mil, EV = R5 015mil. This means that LEW trades at 11.7times the NPV that the assets generate in one year.

JDG’s market cap = R5 980mil, net debt = R670mil, EV = R6 650mil. This means that JDG trades at 6.0times the NPV that the assets generate.

The NPV’s calculated above represent the earnings power of the trading operations for one year. When purchasing these companies an investor buys two assets (1) the trading operation, and (2) the debtors book.

The amount one pays for the operating assets equals the EV less the debtors book. However, in the case of these businesses the debtors book is understated as it does not reflect the full contractual cash flow that flows from prior credit sales. (The instalments that are due include principle, interest and premiums for credit life insurance.)

To determine the remaining contractual cash flows flowing from prior year credit sales one needs to plug these credit sales into an excel model which uses a payment profile to determine when the cash flows will be collected.

LEW: Contractual Cash Flows

Credit sales for the 12months ended Sept 2006, 2007 and 2008 were R1 169mil, R1 293mil and R1 239mil.

Using these credit sales the model predicts that R3 200mil in contractual cash flows flowing from prior credit sales must still be collected by LEW.

Subtracting the R3 200mil from the EV of R5 015mil implies that investors are paying R1 815mil for the operations which are capable of generating an NPV of R427mil per annum. This EV/NPV multiple = 4.2

JDG: Contractual Cash Flows

Credit sales for the 12months ended Sept 2006, 2007 and 2008 were R3 749mil, R3 597mil and R3 061mil.

Using these credit sales the model predicts that R6 900mil in contractual cash flows must still be collected by JDG.

Subtracting the R6 900mil from the EV of R6 650mil implies that investors are paying minus R250mil for the operating franchise which is capable of generating an NPV of R1 111mil per annum.

JDG used to be audited by a medium sized accounting firm who were happy to approve accounts which overstated revenues. The new management team brought in Deloittes to do the audit and have stopped all the accounting fun and games which were used in the past to manufacture earnings. Deloitte’s also forced management to disclose the contingent tax liability. Today, JDG accounts are a fairer reflection of economic reality.

Unfortunately, the same can not be said for LEW where exploiting loopholes to boost earnings remain the name of the game.

LEW continues to recognise revenue upfront. Second, it seems you need to be a double axe murderer to be even considered a bad debt. This is because LEW use recency as the main tool for recognising potential bad debts. The essence of this policy is that if a customer made the last payment his account is thrown into the “good guys” basket against which a provision of 0% is made. Lets say a customer makes his 20^th payment on the due date but has missed 5 other payments along the way. LEW will not provide anything against this customer because the account is recent (the last payment was made) and 75% of all payments have been collected.

In contrast, JDG use the comprehensive method to provide for bad debts. The main focus of this policy is to count the total number of payments that have been missed and to provide accordingly. In the above example, JDG would probably make a provision of 70% of the amount outstanding.

The difference in the provisioning policy is clearly highlighted by contrasting the income statement of both companies over the last 36months.

Against credit sales for the 12months ended Aug 2006, 2007 and 2008 of R3 749mil, R3 597mil and R3 061mil JDG provided and wrote-off R528mil, R825mil and R898mil respectively. This represents 21.6% of the prior 3 years of credit sales.

In contrast, against credit sales of R1 169mil, R1 293mil and R1 239mil over the last 3 years LEW has recorded bad debts of R136mil, R160mil and R190mil respectively. The bad debt expense equals 13.1% of the prior 3 years of credit sales.

The difference in provisioning seems very strange as both companies sell to the same customer from stores across the road from each other. It is difficult to comprehend how JDG’s bad debt experience could be so much worse than LEW’s.

An investor who shorts R4 250 mil of LEW and goes long R4 250mil JDG will effectively be long R138mil cash and own operations which generate R363mil in NPV every year.

Put another way JDG will have to rise from R36 to R67 to be on the same rating as LEW.

1. LEW’s auditors put their foot down – Year End = March 2009
2. LEW investors wake up and start asking questions about why the book is growing faster than sales. (LEW has a large US and EU shareholder base and I have strong reservations whether proper dd can be done when one is so far from the action.)   
3. JDG settle the tax issues with the tax authorities. – March 2009
4. JDG begin to lap a very low earnings base as 2008 earnings are hurt by accounting adjustments and high bad debts as the new management team scrub the book.

Due to much looser accounting standards the quality of LEW’s earnings are not comparable to those of JDG. Alternative valuation techniques show that LEW is in fact twice as expensive as JDG.

This staggering gap in valuation for companies with very similar income streams provides a huge margin safety for this arbitrage trade.   

Catalyst