O'Reilly Auto Parts (NASDAQ:ORLY) is a specialty retailer that sells automotive aftermarket parts and supplies and operates 4,571 stores across the US. Unfortunately, the firm only provides the aggregate number of leased vs. owned stores, which only serves to muddy the perception of the firm's value.
Thus, even though stock is currently trading near a five-year high, the real estate portfolio provides more headroom for further price appreciation. By using a combination of sales comp of similar retail stores and a complete dataset on ORLY's owned stores (from a third party website see methodology for details), I have arrived at an estimate of its real estate holdings of $6.5bn.
Extracting value from real estate is costly and has material implications on the firm's value. Therefore, I constructed a hypothetical sale leaseback transaction that highlights the implication to the firm's cost and capital structure. Even after these changes, the portfolio provides more upside. Land is carried at cost under GAAP. Thus, firms overpay for holding real estate; the financial impacts are capped at a lack of "rent expense" versus comps. The real estate valuation, corrects for this disconnect.
I have derived a share price of $317 by using an EVA valuation methodology and the real estate valuation. My target price implies a 15% upside to the current share price of $275.
I used LoopNet's sales comps for similar stores by state to arrive at the average selling price per square foot by property subtype and geography. Furthermore, LoopNet's database has information on properties by owner and tenant. However, manually parsing the data for 4,700+ individual listings for ORLY to determine which stores were owned and which were leased would be extremely time consuming. Furthermore, the website doesn't allow users to access the database in its entirety; consequently, an iterative extraction process is required to collect the entire dataset. I believe that this was the key deterrent holding back investors from analyzing the data.
Fortunately, I was able to circumvent this issue by writing a web scraping script that extracted information in a few minutes. Learning how to write the script was significantly more time consuming, but the results were well worth the investment. I used the combination of the sales comps and owned store data to conduct a thorough and robust real estate. The total estimated value of its real estate holdings is $6.54bn; however, extracting the value of the real estate portfolio is not a free lunch.
No free lunch: sale-leaseback has material knock-on effects
ORLY will going forward have to foot a heavy rent bill of ~$482m per annum. Its cost of debt will increase from 4.4% to 7.4%, based on a weighted cost of debt calculation. More subtle impacts include holding onto larger stockpiles of cash to meet rent obligations. Therefore, the impact of the sale-leaseback could reduce the pure equity value by -12%. However, the negative impacts are capped and do not increase linearly; its cost of debt will not increase above 8%. Furthermore, a firm cannot record a gain for a sale-leaseback transaction; thus, the resulting liability is also limited.
I used an iterative WACC calculation to determine the appropriate discount rate for my model. Whilst, a fair number of analysts tend to use a static WACC derivation. Given, how significantly the balance sheet warps due to the hypothetical sale-lease back. The impact on share price can differ by as much as 25% versus a static WACC derivation. Furthermore, ignoring the impact on the capital structure is a material violation of steady state assumptions that underpin most valuation methodologies. Consequently, there is no way to circumvent using an iterative WACC.
Forecasting: sales are highly correlated with fleet age
I found that the firm's sales are most sensitive to the age of a light vehicle fleet. The linear regression for this model had an R-Square of 0.93 versus 0.52 for the similar regression run with vehicle miles driven, management's preferred economic driver. Furthermore, the t Stat for both the intercept (-12.73) and beta coefficient (14.66) are highly significant. If the results seem too good to be true, then I hope the following will allay your concerns. Even under the most stringent correction for datamining, Bonferroni's Method (FWER), the revised t Stat would come in at 4.0 given 640 iterations or tests.
The strong statistical relationship between sales and the average fleet age of vehicles also makes intuitive sense; older vehicles are more likely to need regular maintenance and no longer be covered under a warranty program. Moreover, the regression model estimates that top line revenue will come in at $8.56bn, the midpoint of the latest guidance for FY16. Therefore, I believe that the forecasting model is robust. Consequently, I opted to mechanically forecast revenues. Finally, I believe that using a mechanical forecasting methodology that foots with management guidance, better illustrates the value of its real estate holdings.
Sale-leaseback could catalyze share price
The tables above delineate the impact of the sale lease buyback. My valuation model, based off the mechanical forecasting methodology does a solid job of estimating the firm's value, table on the right. It estimates a share price of $278 vs. $275. Therefore, I believe that it's an appropriate model to use for this analysis. That is to say, that since my model's estimate of the firm's value, ignoring its real estate holdings, are in line with the market, then the delta due to sale leaseback can be fully attributed to the real estate valuation.
Therefore, the delta due to sale leaseback can be fully attributed to the real estate valuation. Note that while higher real estate values imply higher share prices, the equity value of the firm takes a hit since the cost of the property determines the implied rent the firm has to pay. However, detrimental impacts are limited; the total value of the lease does not lead to a linear increase in the cost of debt - it's capped at eight percent. In the event of a sale leaseback, the fair value of the firm jumps to $317, a 15% upside to the current share price.
To conclude, I want to reiterate the following: land is carried at cost under GAAP; thus, firms overpay for holding real estate; the financial impacts are capped at a lack of "rent expense" versus comps. The real estate valuation, corrects for this disconnect.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.