Now, that might be a mistake. I might be over-analyzing it and imposing some kind of overly stringent criteria on a bunch of companies that are all cheap. To me, however, there's a big difference between a cash shell like Cadus (KDUS) and a real, cash producing business like ADDvantage (NASDAQ:AEY) that happens to be overcapitalized. I'd much rather own a business with real earnings rather than wait for something to happen with a pile of cash.
My question is this: How cheap is cheap enough? Clearly (to me), George Risk (RSKIA) is cheap at or even just above book value. It's a darn good business so I'm getting high quality assets and earnings power. That gets less clear when looking at lower quality businesses.
Solitron (SODI) sells at 74% of NCAV, has decent z- and f-scores, a FCF margin of 5.3% and an ROA of 12%.
Micropac (MPAD) sells at 83% of NCAV, has similar (slightly better) z- and f-scores, a FCF margin of 6%, but has ROA of 28%.
ADDvantage (NASDAQ:AEY) sells at 95% of NCAV, has similar (in the ballpark) scores and FCF and ROA of 23%.
The slightly better businesses are currently more expensive in terms of price/NCAV. They have less asset-based downside protection, but they are better businesses.
How do you quantify and qualify what is cheap enough? To me, there's a big difference in relative cheapness in a company selling at 74% of NCAV versus one selling at 95%. I'm wondering if I'm putting too much weight on this cheapness measurement instead of acknowledging that any decent business selling at less than NCAV is cheap enough. Yet, one has to have some quantifiable idea of when something is not cheap enough anymore.
Can you help me put this into a unified framework?
There’s a great post over at Oddball Stocks called: “A Stock is a Business”. Read it. Then go over to Richard Beddard’s Interactive Investor Blog. Bookmark that blog. Read it religiously. He looks at Ben Graham type stocks in the U.K. And he looks at them not just as stocks but as pieces of a business.
Here’s what Richard said in a post called “Giving Up on Mastery of the Universe”:
I need to know:
1. Whether the managers have made good decisions in the past, and whether their incentives work in the interests of the owners, because those kind of managers often add value to a company.
2. The products a company sells will still be in demand for years to come, because if they’re not then the past, which we know, does not tell us anything about the future, which we don’t.
3. A company is financially strong enough to withstand the kinds of shocks companies typically experience bearing in mind some are more sensitive to events than others.
4. How to judge whether the share price undervalues the company, bearing in mind the preceding three factors.
That’s it. No. 4 matters. But it matters in just the way he laid it out. This was kind of my point about DreamWorks Animation (NASDAQ:DWA). In that stock, you will either be right or wrong because you are right or wrong about the company. Whether it trades at 0.8 times book value or 1.2 times book value does matter. But it’s not something you need to work out perfectly. You can miss an opportunity because the stock never quite gets below book value. And you can pay too much because you are willing to pay too high a premium for book value. We can’t know intrinsic value precisely. And we can’t value our every investment option against every other investment option with robotic regularity.
So, if someone says simply, “At less than its book value, I’m comfortable buying DreamWorks. At more than book value, I’m just not sure.” I wouldn’t fault them for that. Nobody believes the way the company accounts for its movies can precisely quantify book value in the way you can count cash in a till. So drawing an arbitrary line in the general vicinity of reasonableness is okay. It’s practical. And it keeps you focused on picking the right company. Not trying to elegantly balance price against risk for all stocks at once.
If you believe most decent businesses are worth at least 12 times earnings, you don’t have to drive yourself crazy trying to figure out whether Company A which is superior to Company B is a better buy at 11 times earnings than Company B is at 7 times earnings. If you love Company A, buy it. You will do fine over time if you can buy your favorite businesses at 11 times earnings. This is not an approach that will backfire. And it is an approach you can apply year after year after year.
Now, let’s look at net-nets. Are prices below net current asset value almost always less than the price a private buyer would pay for a company?
Yes. Now, sometimes nobody would be willing to buy 100% of a company. But sellers will not usually be willing to sell a perfectly decent company for less than its net current assets.
That leaves us with a pretty simple arbitrary rule. If a stock sells below its net current asset value it is cheap. If a stock is good and safe and cheap – it will work out over time.
So if your entire investment process consists of nothing more or less than ranking every stock below NCAV from best to worst – the company you’d be most comfortable investing in for the next ten years to the company you wouldn’t hold for a nanosecond – that’s a fine approach. That will work. And it’ll keep you sane.
Yes. You are putting too much weight on the cheapness. If you like SODI, MPAD, and AEY – you should just buy all three.
If you like one of them a lot more than the other two – buy the one you like. Don’t pass on the one you like because it trades for 90 some percent of NCAV instead of 60 some percent.
I’m not saying this because the cheapest net-nets are the worst. They aren’t. I actually have some data to support the idea they may be the best if you stick with the system of buying the cheapest net-nets and holding them through some really tough times. The problem with that approach is that it is based on huge winners offsetting rather frequent losers. And frequent losses are very hard for investors to overlook. So, I’m afraid it’s not a strategy anyone would stick to. High quality net-nets offer a lot more solace even if they don’t result in quite as jaw dropping returns as buying the riskiest net-nets.
Here’s a good way to think about it.
Imagine you did not know what the net current asset value of each stock was – remember, for most stocks no one checks the NCAV. Now, roughly sketch out what you think the normal earning power of each business is (conservatively calculated). Now write down the tangible book value.
Okay, do you have both numbers – the net-net’s “normal” EPS and its tangible book value per share?
Good. Now, I can tell you what stocks generally sell for. I can look at the database of stocks we have here at GuruFocus.
So, what does the average stock trade for today?
The median price to trailing twelve month earnings for all stocks is 16.7. The median price to tangible book value for all stocks is 3.3.
So, if a stock has $6 in tangible book value and $0.50 in earnings per share – that stock would normally trade somewhere between $8.35 a share and $19.80 a share. Let’s call that $8 to $20 a share.
Now, let’s say the stock you are looking at trades for around 85% of tangible book value. You can buy it at $5 a share.
So, the stock you can buy today at $5 a share would sell for somewhere between $8 and $20 if only it was viewed as a median quality company.
Then, why not look for net-nets that aren’t necessarily the cheapest net-net – but are instead the net-net with the greatest chance of one day being viewed as a median quality company.
Don’t worry about how the market sees the stock today. Don’t play a Keynesian beauty contest. In the long run, the stock market is a weighing machine – not a voting machine. We don’t need to poll the market. We don’t even need a scale. We just need to look at these stocks and try to separate the really skinny guys from the really fat guys.
In this case, that means separating the maybe just misunderstood net-nets from the truly terrible net-nets. It means trying to separate bad businesses from perfectly mediocre businesses.
Even if we do that, we are still left with one little problem.
The stock could be worth anywhere from $8 to $20 a share if it was priced like an average company in today’s market. Obviously, if the market’s P/E and P/B multiples contract – you are exposed to the risk that the intrinsic value range I mentioned will fall down to the price you bought the stock at. That could even happen to a net-net if all stocks fell about 40% from here – so the average stock had a P/E of just 10.
I don’t know how to solve this problem. Stocks are worth more in some years than other years. It’s just something we have to live with. In the very long run, you’ll see high P/E years and low P/E years. I don’t know what order they will come in or how long the valleys and plateaus will be.
So it’s always possible for stocks to fall further for a while.
Unfortunately, people tend to focus on the relationship between price and book – and price and earnings – as a sort of acid price test. The least a stock can be worth is – let’s say – 8 times earnings. And the least a stock can be worth is – let’s say – 1 times tangible book.
This isn’t really true. A stock that is expected to have constant losses in the future – but has earnings today – should see no relationship between its current price and its past 12 months of earnings. In reality, there will always be some relationship. Some people buy things like Research in Motion (RIMM) as part of a regular habit of betting that the market gets overly pessimistic when it knocks a company’s P/E down deep into the single digits. They may be right. There are always some value investors like this who buy things purely because they are statistically cheap.
It’s better though if you look at Research in Motion as a business. Now, personally, I have a hard time doing that with a stock like RIMM because the business is not a simple one for me to understand. I don’t understand customer behavior in that space. And my understanding of a business is through its customers. The only other hope of understanding a business is to look at the past. For perfectly competitive businesses this might be useful. Unfortunately, in this case all we can see is that RIMM – like Apple (AAPL) – was basically losing money and then making buckets of money within the same decade. So I have no clue what the past record tells me other than they didn’t have a hot product for a while, then they did, and now it looks like they don’t have a hot product and so – maybe – earnings will vanish again. I don’t know. RIMM is obviously not a stock I can analyze as a business. I can buy it purely on the numbers. But that’s not my thing. So RIMM would be a pass for me.
Now, I’m not saying that all stocks with incredibly volatile earnings are impossible to value on a private owner basis. Let’s look at a Danish company called Egetapper. It makes carpets.
Here is the company’s recent history of returns on invested capital:
From 2004 to 2011 the median return on invested capital was 9%. That’s an after-tax number. I’d take this as my starting point. This is – until we know something different – we are treating this as a business that can earn something like 9% after-tax on its tangible invested assets.
What does this mean in terms of business valuation?
Well, the company could be worth as much as 2 times its tangible book value. With the right use of debt and dividends – we can imagine Egetaepper selling for that kind of price in a stock market that smiles kindly upon Northern European companies and carpet makers. Think of this as being like saying the company could sell for as much as 20 times earnings at some point. Mr. Market might pay that price when he is in a very good mood.
What about the downside?
Obviously, the downside could be 100%. If this is a flawed business competitively it could go under. I’m not writing an article about Egetaepper though so we’ll just make the assumption that this is a company with fairly neutral prospects. It can reliably at least eke out a profit – but it probably won’t grow faster than the overall economy.
Obviously, when you really are analyzing a stock this is the “business analysis” part. This is where you try to figure out what makes a good carpet? How do buyers decide which company to get their carpets from? Where is the best place to manufacture carpets? Does scale matter? How easy is it for others to enter the industry? How easy is it for others to exit the industry? What is the chance of prolonged over capacity? What is the chance people in the future will want nothing but wood beneath their feet? That sort of thing.
In that case, you should look at the price where the rate at which the business’s return on your capital would be sufficient to justify your continued investment. This is much better than asking how cheap can the stock get. The stock could fall to 0.2 times book value for all we know. There is no limit to Mr. Market’s mood swings. But there is a very real limit to what a company can earn on your investment.
In just the past few years, Egetaepper has traded between something like 0.6 times book value and 2.2 times book value. So, sometimes the market thinks Egetaepper’s equity is worth three times as much as what it’s worth at other times.
We can’t let ourselves imagine the lowest market price from the past is some great indicator of the stock’s safety. Instead, we’ll take that 9% return on invested capital and treat it – conservatively I hope – as the return on equity we will receive. This basically assumes Egetaepper will not use leverage.
In that case, if you need a 10% annual return on your investment, you can put the “downside” for Egetaepper at a price-to-book ratio of 0.9. You can afford to hold Egetaepper as long as the company’s price-to-book ratio is below 0.9 when you buy it and its return on equity is 9%. This will give you the 10% return you desire. At prices above 0.9 times book value, you would need to believe Egetaepper could reliably post double-digit returns on equity while you owned the stock.
So, in this case, it happened to work out that book value approximates the highest price at which you can afford to hold the stock. But, let’s look at the key assumptions we made:
· Egetaepper will reliably earn a 9% return on equity
· You will not pay more than 0.9 times book value for Egetaepper
· A 10% a year annual return will be adequate
Think of Egetaepper common stock like a bond. Here, we are buying the book value. But we are buying that book value on the basis of its yield. At about 0.9 times book value Egetaepper becomes a bond with a yield of 10%. At 2 times book value, it becomes a bond with a yield of 5%. You are happy to hold this “bond” as a non-marketable investment as long as it reliably provides a 10% return. Others may one day be happy to buy it from you when it provides only a 5% yield. In that case, you will be happy to turn this “bond” into a trading gain.
Your safety comes from:
· The reliability of the company’s return on your purchase price
· The sufficiency of a 10% return
Now, let’s put aside the fact that if you really did buy Egetaepper you’d need to be worried about Denmark’s currency – something you probably have no opinion on at the moment. You face a similar situation in the U.S. No, you don’t need to worry about exchange rates. But you still need to worry about interest rates. If we head into 1970s style inflation, a 10% return on your investment could go from being enough to give you downside protection to being about the most you could hope for on the upside. In other words, normal P/E ratios today might be 10 to 20. But under severe inflation – and the accompanying very high interest rates – the normal P/E for stocks could be more like 5 to 10.
I don’t have a good answer for how to protect yourself from this kind of risk.
The answer to the other kind of risk is to buy the best businesses you can find – in terms of the reliability with which you expect them to earn your hurdle rate on your investment. Depending on the company, you can look at the hurdle rate as being a necessary return on equity, return on sales, growth rate, etc.
For instance, if you pay 16 times earnings for a stock today, you’re pretty much depending on the company’s growth rate being enough to match your hurdle rate, because it’s unlikely you can sell the stock for much more than 16 times earnings in the future. This changes if the company pays out its earnings. But a company that doesn’t pay out earnings and still trades for a P/E around 16 is a company that needs to give you a growth rate of something like 10% a year just to give you the same kind of protection you get from buying Egetaepper at 0.9 times book value.
So, the way to think about net-nets is to try your best to get away from purely relying on the stock’s price-to-book ratio, price-to-NCAV, etc. and instead to think of the net-net as a junk bond.
Your investment can work out well in a few different ways:
· Someone is one day willing to pay much more for the stock than you paid (like the 0.9x book value to 2x book value Egetaepper example)
· The return on your investment in the stock is sufficient to justify holding the stock (like the 10% ROE possibility for Egetaepper)
· The stock is still worth something close to your original investment after terrible things happen to it
This last bullet point is the one many net-net investors pay attention to. It’s okay to do that. But it’s the equivalent of buying a high yield bond solely on the expectation that it will have more value in default than you paid for it.
That’s a totally legitimate reason for buying something. But the application of that idea is usually kind of narrow. And as a stock investor you aren’t as likely to get the outcome you want simply because an unfettered management team is not going to treat a stockholder’s investment with the same care a court is going to treat a bondholder’s investment. Trust me, I’ve owned many net-nets. And I’ve only been involved in a couple liquidations. Net-nets succeed and net-nets fail. But they rarely liquidate.
So I think it’s important to look at a net-net as not just something that might have value as scrap – but as something that can deliver adequate returns to you while you hold it. And as something that somebody might one day – when dark skies have cleared – pay a “normal” price for.
Look at the reliability of those defenses. How reliable do you think a specific net-net will be in earning an average of about 10% a year on your purchase price? Try to pick the net-nets with the best chance of doing this.
How reliable do you think a specific net-net will be in one day fetching a much, much higher price?
Net-nets vary tremendously on this score. Some almost certainly will trade at prices fairly close to where they are trading now. They have no history of earning high enough returns on capital to justify a price twice what you paid. A company with a 5% return on equity over the last 10 or 20 years, is a company you probably don’t want to pay more than 0.5 times book value for. And, even then, you’d like to have some idea how value is going to be realized sooner rather than later. A 5% ROE stock is not something you want retaining your earnings and growing the business. It’s something you want to flip – the sooner the better.
Other net-nets have very, very high returns on invested tangible assets. These are good businesses. Or at least they were. The future is uncertain. But their balance sheets are clear. These kinds of companies are usually loaded down by cash. This “anti-leverage” is what keeps the company’s ROE so much lower than the return on assets inside its actual business.
A lot of people dislike these companies because they are hoarding cash. And so there is no catalyst. I actually like these companies. You have two defenses in a situation like this. Where ROE is only low because the vast majority of the balance sheet is in cash – you have both the cash and the business. The business is often a pretty good, pretty reliable company all its own. In theory, it is worth a decent multiple of its operating earnings. And the cash is cash.
Now, where there is just cash and no decent business I feel differently. I don’t like those stocks. But I will buy them for the Ben Graham: Net-Net Newsletter if there is nothing good to pick that month. This is basically a way of picking a stock every month but also kind of keeping that money in cash. It’s a cop out. I don’t like it. But I prefer buying a pile of cash to buying a risky operating business.
The worst net-net to be in is a risky net-net. Generally, a risky net-net is something with a low ratio of tangible equity to total liabilities, a business that can’t be insulated in any way from price competition, a product that is quickly becoming obsolete, etc.
For me, a high reliability net-net is one that:
· Has a high ratio of tangible equity to total liabilities
· Has a business that can be insulated from price competition
· And has a product with low risk of obsolescence
Unfortunately, these companies often have high customer concentration. If you’ve seen my picks for the Ben Graham: Net-Net Newsletter, you know I will gladly take high customer concentration if it means I can get a net-net with a rock solid balance sheet, a lasting product, and some pricing power.
How many of these can you find?
Not one a month. I’ve tried.
Maybe one a quarter.
There are just over a dozen net-nets in the Ben Graham: Net-Net Newsletter’s model portfolio. And there are probably 4 stocks that check these 5 boxes:
1. High ratio of tangible equity to total liabilities
2. Business that can be insulated from price competition
3. Product with low risk of obsolescence
4. Ability to reliably earn adequate return on purchase price
5. Possibility to one day trade for a much, much higher price
That means more than 2 out of every 3 net-nets the newsletter bought didn’t meet those 5 criteria. And, remember, I pick the best net-net I can find each month. So, I’m basically saying the best isn’t good enough in 2 out of every 3 months. My hope is that even the two-thirds of net-nets in the portfolio I do not love individually will work out as a group because of their extreme cheapness. But if you wanted to slowly build a net-net portfolio at the rate of one really nice net-net a quarter – I think you’d do fine. With one warning. You have to hold them for a while. Over 3 or 4 years, you can easily end up with more than a dozen net-nets even if you only buy one a quarter. As long as you don’t sell your net-nets too fast.
(Which everybody does. But that’s a topic for another day.)
So, if your question is what price would I be willing to pay for a net-net I thought checked those 5 boxes?
I’d happily pay 99% of NCAV. If your goal is a favorable outcome in as many individual net-nets as possible – rather than having home runs that more than offset your strike outs – don’t look at the price to NCAV. Don’t look at the price to book. Just look at a list of net-nets. And buy the best net-nets on the list.
Yes, it’s better to buy a net-net at 65% of its net current asset value than 95%. But we’ve done both in the newsletter. And, frankly, the 65% of NCAV stock is not even close to being my favorite. I don’t see anything wrong with paying 50% more for one net-net over another net-net, because you are still paying a very, very low price compared to what other stocks sell for.
We are talking about stocks that – if they really are perfectly decent businesses – should be able to double in price even if they start at 100% of NCAV.
To take on a lot of extra risk because you want to hit a triple instead of a double seems odd to me.
That’s why I think it’s worth paying up for the most reliable net-nets. And it’s why I think most net-net investors would do best by starting with a list of all net-nets and then looking for the safest stocks on that list instead of looking for the cheapest stocks on that list.
This is part of a bigger issue with stock selection. Often, it is much better to combine two good traits rather than rely on a single great trait.
People always want to know what are the 10 stocks with the lowest P/E ratio.
Why not ask: What are the stocks with a lower than average P/E ratio and a higher than average ratio of tangible equity to total liabilities?
It’s the same idea. The problem with low P/E stocks is never going to be that you paid 7 times earnings when you should have paid 3 times earnings. It is that you paid 7 times earnings when you shouldn’t have bought the stock at any price.
Same idea with net-nets. It is better to combine cheap enough with safe enough rather than obsess about finding the absolute cheapest stock.
Ask Geoff a Question about Net-Nets
Check out the Ben Graham: Net-Net Newsletter