Description

Haverty Furniture (HVT) is a mid-priced furniture retailer with 120 owned stores (no franchised stores) in the Midwest and Southeast markets. HVT is an undifferentiated player in a difficult industry that is heavily overvalued on an earnings basis. Based on discussions with sell-side analysts and former analysts at two major buyside holders, the market consensus seems to be that HVT’s valuation is attractive on an asset basis. Analysts cite HVT’s market cap of roughly $200M vs. tangible book value of $238M, its lack of debt, and its significant owned real estate assets.

I first discuss why HVT is well overvalued on an earnings basis. I then address where the bull case arguments in favor of HVT go astray. I end with discussing the catalysts that will cause the value gap to close.

The below table shows HVT’s historical operating EBIT for the past 10 years.

This is a business that is clearly in long-range decline, which is logical given trends in the domestic furniture industry and HVT’s lack of clear market positioning/competitive advantage. While HVT does have a brand in its markets, my checks suggest it is a relatively weak brand except for its stronghold in Florida (less than 25% of store base). Furthermore, HVT is getting squeezed on the low end by the growing square footage mass merchants are dedicating to furniture displays and on the high end by specialized boutiques. Fundamentally, it is unclear what reason HVT has to exist in the marketplace and how it is different than the hundreds of mom-and-pop furniture retailers that fade in and out of commercial viability. The best evidence for this is HVT’s failure to earn its cost of capital even during the peak of the housing boom during the mid part of this decade when one would expect furniture sales to be at cyclical highs. For example, in 2005 and 2006, HVT’s book value averaged roughly $275M. This implies 5% ROEs in each year. Compare this to a furniture company with strong brand value like Ethan Allen – ETH earned a nearly 25% ROE in 2006.

Sell-side analysts do not address these arguments, nor do they seem to appreciate the trend in HVT’s long-range financials. 2009 is clearly going to be an awful year for HVT, with 1Q09 operating loss of $7M already equal to the operating loss in all of 2008. Yet, sell-side analysts predict a miraculous earnings turn-around next year with consensus EPS of $0.36 in 2010, $0.37 in 2011, $0.52 in 2012, and $0.63 in 2013. HVT is so overvalued on an earnings/going concern basis that even if we give the sell-side its estimates, the current share price would imply a 16x P/E multiple on 2013 earnings!

What about the argument that HVT has tangible book value of $238M vs. market cap of $212M? Or that HVT has significant owned real estate assets? Is HVT undervalued on an asset basis? 

Taking the tangible book value argument first, yes HVT has a slightly higher tangible book value than its market cap. But due to operating losses, HVT’s net tangible asset value has been falling rapidly from nearly $300M at 12/31/06 to $238M as of 3/31/09. Furthermore, even if the operating losses were to cease, HVT has $278M in aggregate operating lease liabilities which while off-balance sheet must be factored in determining HVT’s true asset value. In fact, in a liquidation or bankruptcy, there would be negative book value - in such a scenario, HVT shareholders won’t enjoy the benefits of HVT’s assets, HVT’s landlords will. 

HVT’s tangible book value of $238M consists on the asset side of essentially PP&E of $188M and inventory of $99M; on the liability side, the driver is $46M of A/P. Of the $188M in PP&E, the footnotes say approximately $100M represents owned buildings and land from about 40 HVT stores. The PP&E footnotes show of the $100M, roughly $50M is land. I view the book value of the buildings as equal to market value because HVT has a 30 year depreciation life on the buildings. The land could have excess value as HVT’s history traces back to the late 1800s, and this is one of the bull arguments on the stock.

However, looking at the earliest 10-K from 1994, land had a carrying value back then of $22M. This means that the majority of the $50M in current book value was added in recent years limiting the amount of appreciation potential. Trends in the land line item in the PP&E footnotes suggest that if we had 1990 or 1985 data available, there would be less than $10M in carrying value at that point.

A more direct approach is I spoke with an ex furniture retail real estate executive at one of HVT’s competitors. He had some intuitive sense for HVT’s store base, and estimated for me than the average parcel was two acres in size. He estimated a max PSF of $20 pad-ready given HVT’s markets and the locations they historically sought. This suggests land value of roughly $80M vs. carrying value of $50M. This fits well with the gains to book value HVT recorded recently on one of its sale-lease back transactions, although a larger sale leaseback transaction HVT closed in 2001 was not done at a premium to book.

The bottom line is that while one cannot value HVT’s real estate precisely without going parcel by parcel through the 40 owned stores, triangulating a variety of approaches does not suggest much more than $30M of upside here from carrying value.

As such, HVT’s asset value does not appear to make HVT cheap at these levels and, when one accounts for the operating leases, HVT could be considered overvalued as an asset play. This leads to a highly asymmetric payoff structure. Because I do not see substantial market value vs. book value differentials in the real estate, tangible book value of $238M is likely the max valuation one can justify on an intellectual basis. For HVT to be worth $238M as a going concern, the company would have to engineer a miraculous turn-around and begin earning its cost of capital after many years of failing to do in spite of the biggest housing boom in 100 years. In this case, the short loses little with a market cap of $200M. For HVT to be worth $238M in bankruptcy/liquidation, its assets would have to worth dramatically more than book value to offset the operating lease liability, which as discussed is not the case.  In either state of the world (going concern or bankruptcy), HVT is severely overvalued. 

Risks:

a)      Short-interest is at 25%. However, the borrow is easy and has low rates.

b)      Excess real estate value – without doing a deeper analysis, it is always possible that there are some real gems in their asset base worth more than the $20 PSF assumption I used.

c)      HVT does a major sale leaseback transaction to generate cash and pays this as a special dividend to shareholders. Alternatively, HVT levers up by mortgaging its real estate base and pays this out in a special dividend. First of all, each of these would be tough to do in this environment in bulk at close to fair value. Secondly, all this is is financial engineering and does not detract from the arguments made above. The only difference is that rather than potentially making 100% on the short, the most you could make is 50% because roughly half the market cap would have been returned to shareholders. Discussions with management indicate that neither option is likely in the immediate future. Furthermore, either course of action would raise HVT’s fixed costs (either greater rent expense or greater interest expense), which when combined with the operating losses I project will only hasten a demise. HVT is much more likely to keep the cash from either action as a cushion against losses than to give it to shareholders. Finally, an attempt to dividend out cash and then file for bankruptcy due to operating losses could well be be treated by the courts as a voidable preference.

d)      HVT gains market share as its competitors which have greater debt burdens fall by the wayside. The sell side makes this argument, but they fail to see that furniture companies have been falling by the wayside for some years now (Levitz, Laz-Y-Boy, etc.) and HVT’s top and bottom lines continue to deteriorate.

Catalyst

The catalyst for this realization is the substantial operating losses that HVT will have to fund over the next few years and in my opinion, indefinitely. As the 10-year EBIT table shows, this business has been on the downward trajectory for a decade and barely broke even in 2007 which was largely a pre-crisis year. Unless we have a dramatic V-shaped recession, HVT will have significant operating losses for potentially several years.

So far, HVT has managed to fund the operating hole through cherry picking its best properties and doing sale-leaseback transactions. However, this does nothing to increase value, in fact it decreases value given that now is an awful time to be selling real estate. Furthermore, HVT has reduced A/R substantially by outsourcing its credit sales incentive program to a third party provider (sacrificing margin in the process). However, there is only $18M in A/R left vs. $60M in 2007 – there is not much room left to go here. As such, HVT will have to fund the extended operating losses I think are likely either through raising equity, raising debt, or continuing to do sale-leaseback transactions of its real estate assets into an awful market. All these courses reduce equity value, and all reduce HVT’s tangible book value which the market seems to be anchoring on. The catalyst will simply be as operating losses continue to occur, HVT’s tangible book value will contract and the share price will contract with it. When the market realizes that this business is not worth nearly 1x tangible book, then the short will truly work but it should be profitable until that point just based on a 1:1 relationship between tangible book and market cap.