Taking Net-Net Seriously

by Geoff Gannon


Over at the Focused Compounding Podcast, Andrew and I just did a 24-minute episode on “investing in net-nets”. It’s episode 17. You can get all my thoughts on net-nets in that podcast episode.

Here, I’d like to focus on just one thought: “taking net-nets seriously.”

There are three lines in that podcast that might really surprise some people. One, is where Andrew asks if George Risk (RSKIA) – then a net-net – was a smaller than normal position for me. I said, “No. It was a 20 or 25 percent position.” Two, I mention I held George Risk for over 6 years. And three, Andrew asked if some of the Japanese net-nets I invested in did really well, some went bust, etc. and I said: “No. They all worked out.”

This is typical of the kind of net-net investing I’ve done. I’ve taken bigger positions in individual net-nets than you might imagine: 5% to 10% position in net-nets where I know next to nothing about the business and 20% to 25% in net-nets where I know and like the business.

And they’ve all tended to work out.

But, please note that I’m using worked out in the sense that you made money, got a decent absolute return, etc. It’s very easy for net-nets to perform relatively poorly versus the market, because they have little to do with stocks generally.

“Work out” here doesn’t mean they beat the market. Often, solid net-nets perform pretty similar regardless of what the S&P 500 is doing. So, what I mean is if you hold a stock for 5 years and it returns 10% a year or more – it “worked out”. It didn’t go bust. But, it wasn’t some sort of home run. Most investors I talk to – and most backtests of purely statistical looks at net-nets – present a very different picture of net-nets. They are risky stocks where huge numbers of them go bankrupt or nearly bankrupt and a few winners become double, quadruples, octuples, etc. just as fast. That hasn’t been my experience. And that’s mostly because I don’t buy into those kind of net-nets.

The best blog about net-nets is Oddball Stocks. The author of that blog once asked me to run some net-net back tests for him. You can tell from the screenshot included in that post that those back tests were run on Portfolio123 (which is my favorite screener). I don’t get paid or anything if you click that link. The reason I’m including a link to Portfoio123 is that people always ask if I like this screener or that screener and the answer is really “No. If I run a screen, it’s one I created on Portfolio123 myself.”

Anyway, here is a post where Nate of Oddball Stocks is talking about more of the kind of net-nets I would normally be looking at:

“…market bottoms offer opportunities to buy ‘quality’ net-nets, companies that have long records of profitability and are selling below NCAV.  There's no reason an investor couldn't combine investment styles, purchasing some quality companies cheap as well as quality net-nets in a low market.

Of course the data shows that net-nets purchased in 2002 and 2009 with long strings of profitability rewarded shareholders.  This is an obvious conclusion, buying almost anything at a market bottom had positive returns.”

What’s good about this post is that Nate then goes through the history of each of those companies to explain why something that looked good at the time really did go on to succeed, or why it somehow failed, and so on. For example, one stock was a fraud. And, he explains that’s why it went to zero.

One thing you’ll notice about the net-nets in that back test is typical of any list of consistently profitable net-nets. You lose money if the company basically goes out of existence. Otherwise, you make at least a little money. Because net-nets are so cheap, buying a historically profitable net-net where some profitability continues on into the future almost always results in you not showing a loss when you sell the stock. Losses of 5%, 10%, 20%, and 40% are pretty common in the big stocks many investors buy (this is because P/E multiple contraction is pretty common). Those kinds of mild losses are not common with net-nets chosen on the basis of past profitability. You tend to see either a 100% loss or some gains over time. Now, if you hold the net-nets too long – and especially if you hold them during a bull market – you may underperform the S&P 500. But, you’re unlikely to experience any sort of medium sized losses. In fact, that’s why I encourage people to always focus on the downside with net-nets and never the upside. If you could somehow know that the net-net you’re buying will be profitable in every year you own it – odds are the stock will do well for you. In fact, the odds are overwhelmingly in your favor (because you’re buying the stock so cheap) that just clearing that modest hurdle of “staying profitable” is all you need out of your net-net selections.

In fact, in the 2002 back test in that post at Oddball Stocks you basically had 8 net-nets where the investor wouldn’t lose any money in the stock and 2 net-nets where they’d lose everything. One of the two total losses was a fraud.

The other thing that’s notable about net-nets is that they often stay net-nets – or stay below book value – permanently or intermittently even when they perform as well as the S&P 500. This is especially true of small, illiquid, overcapitalized net-nets. A lot of people who come across a net-net will ask: “Can this stock just stay a net-net forever?”. Sometimes, it can. There are examples of stocks that are almost always net-nets, near net-nets, etc. like George Risk and Micropac Industries that constantly show up on net-net lists and yet increase their stock price over 5, 10, 15, and 20 years at pretty normal rates for a public company. This, of course, means you risk being stuck in a stock that is cheap when you buy it and cheap when you sell it and only returns 8% to 10% a year while you hold it. That’s a very real risk in net-nets. But, earning 8% a year long-term in most assets other than net-nets is just considered a normal outcome.

So, why do investors feel so bad when they buy a net-net and it is “dead money” in the sense it only returns 8% to 10% a year over the 5 or more years while they hold the stock?

The biggest reason is that I think they realize they’re making a special effort to dig up a net-net and imagining the riches that will come from discovering some previously undiscovered stock. So, if Apple returns 8% or 10% a year from here – that’s understandable. You aren’t taking unusual risks, putting in unusual effort, etc. You didn’t buy it thinking it was dirt cheap. But, with a net-net you did. So, you’re waiting for something to happen with the net-net. And what you mean by “something happening” is not chugging along at 8% a year.

So, investors remember their experiences in net-nets differently. They often remember a net-net didn’t do much – but that’s because they’re getting frustrated that the value gap isn’t closing the way they imagined. They imagined making 50% in one year. Instead, they made 50% over 5 years. Of course, that’s par for the course in stock investing generally (and the net-net is still cheap when they sell out). But, you probably go into big caps expecting to hold them as they chug along and yet you go into net-nets expecting to get your “one puff” from the cigar butt.

Why else might investors feel “dead money” experiences in net-nets are more frustrating than getting the same annual returns in bigger stocks.

One, some net-nets have very low volatility. I mean very, very low volatility. I can’t emphasize this enough. It’s something most people aren’t used to at all. I’m used to the experience of ultra-low volatility in some stocks from owning them and from back testing them. If you build a net-net portfolio that matches the market over a 5, 10, 15, or 20 year back test – it’s going to do it with a lot less volatility than the market. I don’t talk about beta on this blog. I don’t think about beta. But, the truth is that any net-net portfolio built on the kind of criteria I care about: long history of profitability, high current assets (especially cash) versus total liabilities, etc. is going to have a very low beta. At the high end, the beta might be about what you’d expect in super defensive mega caps. At the low end, it’s unimaginably low. It’s a number stock pickers aren’t used to ever seeing in a portfolio.

But, I think this is incredibly misleading.

Because, it’s not like low-beta giant stocks. Giant, low beta stocks move with the market they just do it more gently. A net-net portfolio isn’t moving with the stock market at all. There’s some correlation. But, we can find assets that aren’t stocks that are at least as highly correlated with something like the S&P 500. And it may seem odd I’m mentioning the S&P 500 when talking about net-nets. That’s not the right benchmark to use, is it?

Well, the personal portfolios investors who write to me talk about look an awful lot like the S&P 500. So, if they carve out 25% of their portfolio to be a dedicated net-net portfolio – it’s going to look very strange in terms of the red and green arrows they see on their account page. Most investors are used to going into their account on a day when the S&P 500 is up 2% and seeing a sea of green and when the S&P 500 is down 2%, they see a sea of red. The net-nets in your portfolio won’t follow that pattern. That isn’t because they’re net-nets. It’s just because the buying and selling of these stocks is not in any way driven by people’s attitudes about stocks generally.

Net-nets are also usually small stocks. There’s very little news about them. Many consistently profitable net-nets don’t change all that much from year-to-year.

These are the only explanations I have for why someone with a big cap stock that returns 8% to 10% a year over the 5 years they own the stock tends to think that stock wasn’t dead money and yet a net-net that returned 8% to 10% a year over 5 years was dead money. From a news perspective, the net-net was quiet. And from a stock price perspective, it didn’t wiggle up and down that much. Big cap stocks give a much greater sense of action than the safer, higher quality net-nets I advocate seeking out.

I’ve also noticed that any stock where you have to use limit orders and wait hours or days or weeks to get your order filled is seen differently by investors. An illiquid stock literally doesn’t move some days. Now, month-to-month it always moves. So, this shouldn’t matter much (when was the last time you sold a stock within a month of buying it?). But, the presence of day-to-day movements seems to create more of a feeling of movement in a stock even when the month-to-month movement is the same in a liquid stock and an illiquid stock. People check their portfolios more than once a month.

Of course, some net-nets are plenty volatile.

Why haven’t I talked about these low quality, unsafe, not historically profitable net-nets?

I don’t invest in them. So, it doesn’t matter to me if those net-nets make up 90% of all the situations out there. I’m not blindly picking net-nets. I’m not buying a net-net index. I only buy certain kinds of net-nets. And I’ve had enough experience with those net-nets to know they don’t behave like a purely random list of net-nets screened for on the web.

What does a random net-net screen look like?

Some of the companies are frauds. You are going to see a surprising amount of U.S. listed net-nets that are really companies controlled by Chinese citizens doing business in China. Don’t touch those. You’re also going to see a surprising number of companies incorporated in Nevada. I’m not saying those are frauds. But, I am saying that if you pay close attention to the stocks already in your portfolio – those will tend to be incorporated either in Delaware or in the state the company has long been doing business (where it was founded). If you compare your current portfolio to a blind net-net screen, I think you’ll find the results of that screen will have a greater percentage of companies incorporated in Nevada than your portfolio does.

You’re also going to find what I call “data errors”. Some of these are caused by the actual data the website is pulling being bad. But other times, it’s really more like computers lacking human common sense. So, a computer may tell you a homebuilder is a net-net. It’s not. I still think it’s interesting when a homebuilder you know shows up on a net-net screen. But, it’s really not a net-net. If you do find a homebuilder selling for less than its inventory of land (held at cost) less its total liabilities and you know where that land is and how good it is and so on – that could be a good opportunity. But, it’s similar to seeing America’s Car-Mart (CRMT) on a net-net screen. When I wrote a report on Car-Mart, I said it should be valued based on receivables. So, stock price divided by receivables per share is a good way to check how cheap the business is. But, those receivables are very, very low quality debt. They are stated on the balance sheet at well below the amount actually borrowed by the car buyer. Again, America’s Car-Mart can be a really interesting stock when its share price falls below its receivables per share. But, you have to understand the business very well to make a judgment about that. You need to think about where we are in the auto loan cycle, how many months are left on the average loan, what incentives are for the employees who make and collect loans at Car-Mart, whether management is candid, what Car-Mart’s charge-offs have been historically, etc.

A human would see that almost all the net current asset value of a homebuilder is land and almost all the net current asset value of a used car seller is risky loans. A computer won’t notice that.  

I’m always saying you should look for net-nets with a long history of profitable and a lack of liabilities. That means I’m saying you need to cross off about 90% of any net-net screen results you get.

So, when people ask me why I don’t seem to be buying any net-nets lately, it’s because there’s always a lower supply of the kind of net-nets I like. But, remember I had 20% to 25% of my portfolio in net-nets for the 6 years I owned George Risk, I had 50% of my portfolio in net-nets when I held Japanese net-nets, and I just wrote about Pendrell (PCOA) – a stock I don’t own as of now – within the past two weeks.

So, I do invest in net-nets. I’m not done with buying net-nets. I’d like to buy more net-nets sometime in the future. I’m just selective about what net-nets I buy. And it’s because I take net-net investing as seriously as any other kind of investing I do. I’m looking for the same margin of safety, history of profitability, strong solvency, etc. that I look for in billion dollar market cap stocks I buy.

Finally: the question everyone asks?

Does buying net-nets with low liabilities and a long history of profitability really outperform blindly buying all net-nets. The honest answer is that I don’t care. I’m not an empirical investor. If a back test works, but I don’t feel confident in the logical principles on which the back test’s strategy is based – I’m not going to consider that strategy. It may be that buying the most leveraged net-nets that aren’t making money right now is the best approach because a few big winners offset the losers. But, I wouldn’t know how to evaluate that strategy.

Ben Graham said investing is most intelligent when it is most businesslike. I don’t know how to buy more leveraged, less historically profitable net-nets in a businesslike way. Seven years ago, I bought George Risk not just because it was a net-net, but because it seemed like a businesslike investment. If I could buy the whole business for less than net cash – I would’ve. So, why not buy a piece of it in the stock market.

Listen to the Focused Compounding Podcast

Visit Oddball Stock