Is a sum of the parts valuation worthless?

In the aftermath of catastrophic losses on Horsehead Holdings (ZINC) a number of value investors have been asking themselves what went wrong.  Some have said the culprit has been the use of a sum of the parts valuation.

Horsehead Holdings is a Pittsburgh, PA based company that specializes in zinc oxide processing.  Famed value investor Monish Pabrai found the investment during the depths of the financial crisis trading for less than net current asset value.  As the economy, and the company's shares recovered Pabrai continued to talk about the investment, and it transformed from a deep value playt to one with a great management team and an attractive story.  The company supposedly had a best in class processing system that allowed them to process zinc cheaper than peers and as a result once their factory was operational they'd mop the floor with their competition.

Unfortunately the company hit a perfect storm.  Zinc prices declined and at the same time the company's next generation system experienced continual failures and needed of costly repairs.  To make matters worse the company had a risky balance sheet loaded with debt expiring in the near term.  These factors collided and the company now trades for mere pennies down from $15.18 earlier this year.  They missed a debt payment and will most likely be entering bankruptcy soon.

One narrative I've heard a few times out of the investment is that a sum of the parts valuation methodology is dead, or worthless.  A sum of the parts is where an investment is valued on individual pieces of the company rather than the company as a whole.  For Horsehead their next generation plant had a market value, along with two other ancillary businesses they owned.  Investors looked at this situation as good downside protection because in theory management could sell the side businesses to fund the construction issues.

I don't believe the sum of the parts valuation methodology is worthless, but I do believe it's misapplied.  I want to explore applications and uses in this article.

Let's consider a very simple example.  Imagine you had a bike, a shoulder bag, a cell phone, and nice rain coat.  What is it worth?  This is a simple question, you could find the value for each item separately online.  With eBay or Craigslist you could probably find each exact item in the same condition and determine within a few dollars the true market value for this entire collection.  If you were to go through this exercise and determine the collection is worth $1,500 and someone offered you $1,000 to buy the ensemble you'd probably think "that's crazy, I can sell each item individually for more than $1,000."  That is the essence of a sum of the parts valuation.

But take a step back, if you have all of those items you also have all of the pieces needed to be a bicycle messenger.  And maybe that's why you do have all of those pieces, it's because you make deliveries on your bike.  Now the situation is different, by using all of these items you might earn $10/hr, or $1,600 per month.  If someone were to offer you $1,000 for the ensemble you'd laugh because by putting the pieces together and using them to generate an income would result in $1,600 per month until you decide to stop riding.

The second example is why in most cases a sum of the parts makes no sense.  A company's parts, their factories, their land holdings, their machinery, any other assets are much more valuable as an operating entity as opposed to being parceled off.  And the reason all of those assets are together in the first place is because at some point someone tried to make a go of being an operating entity.  The mandate for those assets is to work together and to create something, that is their purpose.

A sum of the parts makes sense in two situations, the first is if the company has excess assets that on their own are extremely valuable.  Consider our messenger example, what if I said you also owned a set of rare historic bikes in addition to the messenger setup?  These bikes are stored in your garage and never used for deliveries.  If you were not emotionally attached to those bikes you could potentially sell them for thousands apiece without affecting your messenger job.

The second situation when a sum of the parts could be appropriate is when management has made it clear they are willing to part with significant portions of their operations in an effort to streamline/refocus/monetize their company.

If management views their assets as available for sale, and actually sells them when given an attractive offer then it's reasonable for investors to view them as the same.

The key to a sum of the parts analysis is that management needs to be willing to monetize their assets.  A company with an extremely valuable piece of land that they are unwilling to sell because it was purchased by the CEO's great grandfather has a value thats determined by the company's use of it.  Management is the key to unlocking value at companies with valuable assets.

In an issue of the Oddball Stocks Newsletter I wrote about Du Art Films, a film processing company located in New York.  The company owns an extremely valuable plot of land in Manhattan that is underutilized.  They could sell their land to a developer to build a skyscraper for many times their market cap.  They own other valuable assets as well.  The reason shares trade so cheaply is that management has been unwilling to monetize those assets.  But there are indications this could be changing as management is quite advanced in age, and shareholders are agitating for change.  But without management being willing to do anything it's likely these assets will remain undervalued for a while.  Du Art is a typical sum of the parts situation, a set of valuable assets that are at the whim of management.

If a company has excess salable assets and a management that appears willing to sell them then a sum of the parts could be appropriate.  If a company is simply a collection of unrelated assets and there is a management team willing to monetize them then a sum of the parts could be appropriate.  If a company has entrenched management that has a "down with the ship" mentality a sum of the parts valuation definitely isn't appropriate.

In general it's best to look at a potential investment through a number of different valuation lenses.  If a company comes up cheap when using multiple investment approaches then there's a very strong chance that it's truly cheap overall.  But if a company appears expensive using every approach except one, and that one approach says it's blindingly cheap it's probably best to look deeper or look elsewhere.

Like anything, a sum of the parts valuation methodology is appropriate in some situations, but no appropriate in all situations across the board.

11 comments:

  1. This is an excellent post and I'm glad you wrote about it, because it can be applied in many scenarios including analysis of net-nets.

    I don't know the details of Du Art Films, but a caveat to applying the sum of parts analysis is that you have to be able to see it through, or count on someone else to do it. If you have the capital, and the market will sell you the amount needed, then you can buy a controlling stake and make it all happen yourself. Otherwise, you have to count on someone else doing it or management changing their mind, because like in your example, they may have a passion for being a bike messenger and it can become tough to make good financial decisions when there are other considerations like a person's own passion/desires. (Why else would diamonds have jewelry value?)

    Anyway - great post, keep it up.

    ReplyDelete
  2. I like your conclusion, especially about back up systems/triangulating value.

    However, I disagree with your initial example. The example of selling your bicycle messenger equipment is more akin to replacement cost analysis than to sum of the parts. Typically, SOTP is applied to different operating units/entities (except in real estate spin offs which seem to be in vogue right now). These businesses are able to operate and generate income on there own. I don't believe your example works b/c your bike, cell phone, bag, etc. all should have separate income streams to really apply SOTP effectively.

    I also disagree with saying they are worth 1.6k/month b/c that ignores the what the person doing the work is worth or at least the opportunity cost of their time.

    That being said, aside from my nitpicking on your example, I think the conclusion you reach is entirely valid and your article is thought provoking.

    ReplyDelete
  3. An excellent post Nate. I'd like to add:

    Sum of the Parts is a great mechanical tool. But its all about context. Management attitude is one consideration, the other is time. Time is just as important because the longer the value sits unrealized, the less it may be worth. Its not just inflation that eats value, its change in market conditions. You see, valuing assets at market v book is heavily dependent on an informed market of sufficient depth. When cycles change, technology changes, or debt becomes due, time is not the friend of value.

    Generally, property is the best sum of the part asset to get value out of- even though of course, real estate valuation is still quite prone to interest rates and liquidity (debt availability). However, technology and inflation are generally not a problem. A related The problem to time, is quantity. If there is too much to liquidate, discounts are expected and sometimes if the meal is too big for the market to swallow, the discount on what otherwise seems like market value will evaporate with transaction costs.

    ReplyDelete
  4. An excellent post Nate. I'd like to add:

    Sum of the Parts is a great mechanical tool. But its all about context. Management attitude is one consideration, the other is time. Time is just as important because the longer the value sits unrealized, the less it may be worth. Its not just inflation that eats value, its change in market conditions. You see, valuing assets at market v book is heavily dependent on an informed market of sufficient depth. When cycles change, technology changes, or debt becomes due, time is not the friend of value.

    Generally, property is the best sum of the part asset to get value out of- even though of course, real estate valuation is still quite prone to interest rates and liquidity (debt availability). However, technology and inflation are generally not a problem. A related The problem to time, is quantity. If there is too much to liquidate, discounts are expected and sometimes if the meal is too big for the market to swallow, the discount on what otherwise seems like market value will evaporate with transaction costs.

    ReplyDelete
  5. An excellent post Nate. I'd like to add:

    Sum of the Parts is a great mechanical tool. But its all about context. Management attitude is one consideration, the other is time. Time is just as important because the longer the value sits unrealized, the less it may be worth. Its not just inflation that eats value, its change in market conditions. You see, valuing assets at market v book is heavily dependent on an informed market of sufficient depth. When cycles change, technology changes, or debt becomes due, time is not the friend of value.

    Generally, property is the best sum of the part asset to get value out of- even though of course, real estate valuation is still quite prone to interest rates and liquidity (debt availability). However, technology and inflation are generally not a problem. A related The problem to time, is quantity. If there is too much to liquidate, discounts are expected and sometimes if the meal is too big for the market to swallow, the discount on what otherwise seems like market value will evaporate with transaction costs.

    ReplyDelete
  6. I haven't looked into Horsehead in detail, but I guess one of the problems seems to have been their full recourse debt. Sum of parts valuation is not very helpful if the company is in risk of default.

    ReplyDelete
  7. What an absolute disaster for Mohnish. On something that you could clearly see coming. And boiling down his thesis--it evolved into taking management's word that the Mooresboro plant would generate $100 million in EBITDA at completion. Management that had chronically over-promised and under-delivered. Built a plant with 40+% cost overruns. Went from a great balance sheet to raising capital via dilutive equity. Consistently overestimated operating capacity after constant delays (you could count on it, literally every quarter). And he had the nerve to commend management and say he didn't learn anything from making the bet. Nothing about these guys deserved commending. And he was the largest holder and set by idly while the company took bet the company risk and failed. What an embarrassment to him and a slap in the face to his investors.

    ReplyDelete
  8. Good question and article Nate.

    When looking at sum-of-parts valuations, I fall back on Walter Schloss' approach since he clearly mastered this area. If we think of a low price-to-book value as a proxy for a sum-of-parts valuation, then Schloss' other criteria are listed below. The obvious box that didn't get ticked for ZINC is long-term debt.

    http://www.aaii.com/journal/article/finding-value-amoung-the-lows-the-walter-j-schloss-approach.pdf
    Criteria for Initial Consideration
    • Ten-year track record
    • No long-term debt
    • A low price-to-book-value ratio
    • A stock at or near its 52-week low price
    • High insider ownership

    Similarly, Seth Klarman also lists 3 valuation models in his book - Margin of Safety - including a liquidation model which provides a sum-of-parts approach. He focuses on risk reduction around debt to ensure return of invested capital and also portfolio position sizing within the stock purchase decision.

    Mohnish Pabrai is a tremendous investor, but like all of us have at some point, he had an own-goal with ZINC. I could list my own failures, but let's stop this long comment here.

    - Phil

    ReplyDelete
  9. Mohnish has now under-performed for 13 years (with significantly more concentration/risk, regardless of what he may say), so the halo is wearing off fast and this includes 03 and 04, when he racked up against the indices. Over the last ten years, he is off massively against every benchmark. And this is through 12.31, so before the Horsehead zero. He make need to re-read some of his books on cognitive biases, because he is quickly becoming a living example of the Dunning-Kruger effect. Someone putting 10%+ of the bankroll into single names/50% into single industries had better be careful after so many zeros, or he risks turning into another case study.

    ReplyDelete