Get Premium to unlock powerful stock data

Backtesting Net-Nets: Does It Matter?

Author's Avatar
Geoff Gannon
Apr 02, 2012
Article's Main Image
Someone who reads my articles asked me this question:

Hi Geoff,

Quick question. There was an old blog post at Oddball Stocks. He made up a rough Net-Net index for one year and noticed that it beat a global all-stock index by a significant margin. He noted that the performance difference came from relatively very few stocks and that the rise in prices of some of those stocks would probably not be attainable by an individual investor as one would likely sell after the position quadruples instead of holding until it octuples (made that word up) and other such potential performance barriers. And a person that picked net-nets individually is likely to not pick the stocks with monster gains. The GuruFocus Net Net newsletter creates a basket. I understand you personally are more focused. Is the research you've found on historical net-nets results meritorious of stock picking and focusing net-nets instead of basket/index strategies? You've mentioned high insider ownership and smaller market cap are favourable for strong results with net-net investing but is that quantifiably backed up with exhaustive backtesting and in a focused portfolio over a net-net index?

Thanks as always,


Do you mean Cheap Stocks? I don’t remember an index at Oddball Stocks. But I might have missed it.

No. The newsletter’s approach is not backed up by backtesting versus a net-net index. In fact, huge pools of every available net-net actually do really, really well in backtests. It would be quite a chore buying them all. But that’s another story.

The idea of the Ben Graham: Net-Net Newsletter is not to beat a net-net index. There is no need to do that. If you buy a basket of all the net-nets you can find and you hold them for a set time period – without regard to their price increase – you will do very, very well over time. The problem is that investors will not do this. They will sell their net-nets too soon. And they will not realize these gains.

For example, if you just bought all the net-nets available on this day – April 2 – in each year from 2002 through 2011 and held the net-nets for one year – no matter how their prices changed during that year – and then sold only those stocks that no longer qualified as net-net and then repeated the process each year you would beat the market in 8 of the last 10 years. And you’d end up with a compound annual growth rate of more than 27% from 2002 through today. Anyone would be happy with that result.

Is it attainable?

Look, nobody would carry out that exact process. It would require buying lots of small stocks. While we talk about how few net-nets there are – the truth is that if you literally bought every net-net every year you could fill an individual investor’s portfolio with no problem. In fact, at some times, an individual investor would be overwhelmed by the sheer number of net-nets they were supposed to buy if they followed a strategy of buying every net-net.

To give you some idea – I’ll use today’s date (April 2) as the day we will pretend we bought every net-net for the last 10 years.

How many stocks would we have to buy?

The low – which we hit in April 2006 – would be just 40 net-nets. Half of these 40 net-nets would be holdovers from the year before. In general, there is about a one in two chance that a net-net will still be eligible for inclusion in an all net-net portfolio one year from the day you first buy it. This varies from year to year. Last year was – relative to stocks generally – probably the worst year for net-nets in at least a decade. There is no hard data on this. But it was an unusually awful year for net-nets.

So does buying a basket of every net-net mean just buying around 20 new stocks a year?

No. In April of 2009, it would’ve meant owning 358 stocks – of which a whopping 288 would be new positions. Net-nets are a symptom of the overall stock market’s health.

Why were their hundreds of U.S. net-nets in 2009?

Because the stock market crashed. Prices fell without regard to liquidation value. Or value of any kind. Basically, any stock could fall 30% to 60% in late 2008 and early 2009 without regard to whether it was a good company, had already been trading at near its liquidation value, etc.

In Japan there are net-nets for a different reason. Many Japanese net-nets have increased their net current asset position – they have increased cash, inventory and receivables and decreased liabilities – while their stock prices have fallen. The process has not been very fast. In some cases, you can plot the transition over a 10 year period. The stock price generally stays flat or falls for 10 years. Meanwhile, the net current assets per share generally stays flat or rise for 10 years. At some point they cross. Then they just keep going their separate ways.

So, in a market like Japan – which has been in decline for a long time – net-nets are relatively common at all times. In the U.S., net-nets are common only when the stock market drops sharply.

For example, using our April 2 date as the day on which we measure the number of net-nets each year, we’d find 200 to 300 net-nets in 2002 and 2003. Then just 40 to 50 net-nets from 2004 through 2007. Then 140 net-nets in 2008. The aforementioned 358 net-nets in 2009. And then anywhere from 70 to 100 net-nets in April 2010, April 2011, and today.

That’s a lot of net-nets. For the last three years, it would mean buying 35 to 50 new (very) small stocks every year. That’s not something individual investors do.

And they definitely do not hold their net-nets for a year – regardless of price increases.

In fact, this is one of the complaints I get about the Ben Graham: Net-Net Newsletter. Isn’t waiting one year between checking up on a stock too long?

No. It’s probably too short. Most investors would be better served by ignoring a net-net for two years after they buy it. If a net-net goes bankrupt within two years of you buying it – your mistake was buying it in the first place, not forgetting to sell it. You can look at the stock’s leverage, business risks, Z-Score, and F-Score when you buy it. If a stock that scores well on those measures fails within two years, that’s a very unusual situation. As part of a portfolio of dozens of net-nets, it’s not something I’d worry about. It’s a fluke.

Now, in reality, many people will buy net-nets without regard to their safety. And many net-nets that people buy have very bad F-Scores. Some also have very high leverage in the sense that net current assets divided by total liabilities is not very high. I bought a stock like this last month for the Ben Graham Net-Net: Newsletter.

It might pay off. On a risk-return basis – a pure handicapping approach – it may be a good choice. But it’s very risky. It could certainly fail in the next two years. It’s in an industry where failure is always possible.

I made the riskiness of this stock clear from the opening words of last month’s newsletter:

“(March’s pick) is a “cigar butt” with a weak balance sheet and volatile earnings. It is too risky to be a standalone stock pick. But at a 33% discount to its net current assets, (it) is an acceptable addition to a diversified net-net portfolio.”

If you buy 30 stocks exactly like our March pick, you will get good returns over time – as long as you do not sell your winners too fast. You can’t sell a net-net that rises 30% in three months just because you would like to take your profit and get out of the stock. I mean, you can – but you shouldn’t. Your result over years and years of net-net investing is going to be much, much worse if you sell your net-nets the moment they start rising. That is a sign that you don’t really have the stomach necessary to own net-nets. If your first thought is to sell a net-net the second it rises – you shouldn’t buy net-nets. Remember, your winners have to offset your losers. And you will have losers. The reason a stock becomes a net-net is because just about everyone who looks at the stock hates it, thinks it is a dinosaur, a fraud, an imminent bankruptcy. The collective wisdom of the crowd many not always be right – in fact, in the aggregate it is always very, very wrong when it comes to net-nets – but you would think the consensus opinion might once in a while end up reflecting reality. In the cases where it does reflect reality, a total loss is possible.

Most people who buy net-nets will lose money in net-nets. They aren’t going to hold their winners long enough.

I don’t want to make this sound like I’m advocating a momentum approach. Momentum and value do both work in the stock market. I don’t pay any attention to momentum. But I’ll tell you this about net-nets – and other deep value stocks – they have a really long runway.

It’s like buying a great growth business at 12 times earnings. When it rises to 15 times earnings, you keep it. At 20 times earnings – if this is Starbucks or something, why not keep it? Maybe we could pick a number – 30 or 40 or 50 times earnings – where you should sell a growth stock no matter how great its next decade or two of expansion will be.

I don’t know. I’m not a growth investor. But I can do the math. And if you’re absolutely certain you have one of the greatest growth companies on the planet in your portfolio – the right price to sell at is never. It’s not something to think about. The time you sell at might be. But a great growth company is not a stock you want to trade into and then out of without participating in the growth you bought it for.

A net-net is like that. Only it’s for a different reason. And it’s more compressed in time. But a net-net has a really long runway too. Because a net-net is a business that’s been left for dead. A living, breathing business is worth a lot more than net-nets trade for. So once it becomes obvious to investors that this net-net has life in it – it’s not like buying a blue chip stock at 12 times earnings and selling at 18 times earnings.

There is no relatively tight and agreed upon band of reasonable prices for a net-net. There is failure and death on one side. And there is being valued like a living, breathing business with earnings and a P/E ratio and people talking about your future on the other side.

If investors go from thinking a net-net is a bet for or against bankruptcy to thinking a net-net is now a legitimate business that should be valued like the thousands and thousands of other stocks out there – the relative change in price will look huge to you.

So what?

It was probably pretty huge on the downside to. You just weren’t there for the years when the stock feel from $48 to $8. But, suddenly, if you get in at $8 and the stock rises to $16 – you start getting dizzy. How can this dog be up 100%? You’ve got to get out.

Well, it was once $48 and it was once $8. The market was probably wrong on both prices. I’m not saying the range of past prices is the best guide to where a net-net should trade.

But I would suggest checking the very long-term stock price history on a net-net. You’ll often find that even after a 100% increase, the stock is still less than 50% of its all-time peak. The stocked I pick for the Ben Graham: Net-Net Newsletter last month is one such example. If it doubled from where we bought the stock, it would still be trading at less than one-third of its all-time high price. Now, yes, the all-time high was a wacky price. But so was the all-time low. So was the price we paid. A net-net can rise 30% or 50% or 100% or 200% and still not be the least bit expensive.

Let it.

You mentioned high insider ownership and low market cap as being signs of a good net-net. And as being something I look for in a net-net. That’s partially true. But I need to clarify. Because it’s the big picture impression that matters here – not the specific numbers we’re looking for.

The issue here is simply the “promotionalness” of the company. Low market cap isn’t a better measure than simply “Shares Outstanding.” In fact, the best way to define a net-net that is “non-promotional” (a neglected rather than hated net-net) is to simply count the number of shares owned by institutions.

By the way, this is true of all stocks. A very pure measure of how much Wall Street knows about a company is usually the percentage of shares owned by institutions times number of shares outstanding. By this standard, companies like Seaboard (SEB) are shown be a mystery to Wall Street (true) rather than being assumed to be a well-known company simply because they have a $2 billion market cap. Actually, the company is much more of a mystery than some companies one-fifth its size.

When you start looking at how stocks with the same metrics like P/E, PB, P/S, being a net-net, having a fast growth rate, etc., do compared to each other – you notice something. Measures like institutional ownership, analyst coverage, short interest and shares outstanding do a good job of sorting seemingly similar companies into long-term winners and losers.

What those measures really do is just one thing: They measure how much Wall Street knows about a company. If a stock has low institutional ownership, low short interest, no analyst coverage, and a small number of shares outstanding – nobody is looking at the company.

In general, it is best to buy net-nets where institutions own the fewest shares. I don’t mean (just) the lowest percentage of shares. I mean literally the fewest shares.

Here’s an illustration.

I used this simple measure – absolute number of shares owned by institutions – to sort net-nets from least known to best known.

Then I created a list of the 13 least well-known net-nets for each year (as of April 2 of every year).

I went back and looked at the 13 net-nets where institutions owned the fewest shares on April 2, 2010 (exactly two years ago). They were:

· Solitron Devices (SODI)

· Wireless Xcessories Group (WIRX)

· Empire Resources (ERSO)


· JLM Couture (JLMC)

· Soapstone Networks (SOAP)

· International Baler (IBAL)


· Electronics Systems Technology

· Micropac Industries (MPAD)

· Impreso (ZCOM)

· Merisel (MSEL)

· XO Holdings

Ten of the 13 stocks are up over the past two years. Remember, I suggested there’s no real reason – if you’re buying a basket of net-nets – to check whether you should sell a net-net you own within two years of buying it. Let the price do what it does. If it runs up – let it run up. It was a super cheap stock to start with. It can double or triple and still be cheap. Don’t sell a net-net just because it goes up.

The best performers in that group were super tiny stocks: International Baler (IBAL), ELXSI (ELXS) and Empire Resources (ERSO).

The worst were Electronic Systems Technology (down 19%), Wireless Xcessories (down 22%) and Soapstone (total loss!).

The median return was 68% over two years. Or about 30% a year (compounded). So, at least over the last two years, it certainly didn’t require owning a few big winners to make money in net-nets. If you looked for the least well known net-nets, bought the top 13 most neglected stocks, and then didn’t touch them for two years – your median result was a 30% a year gain.

This is both a blessing and a curse. I think most people would sell their net-nets before they ever saw a nearly 70% gain over two years.

Some of the stocks on that list don’t file with the SEC. As a result, I don’t pick them for the Ben Graham: Net-Net Newsletter. For my own portfolio, I would go ahead and buy companies that don’t file with the SEC. The SEC provides a convenient place to gather all a company’s financial material (EDGAR). Otherwise, I don’t actually care whether a company files with the SEC or not. Many companies that file with the SEC are frauds. Many companies that do not file with the SEC are not frauds.

Even the companies that don’t file with the SEC did once file with the SEC. And you can find out why they stopped filing by reading their reports just before they stopped filing.

Regardless, I don’t pick stocks that don’t file with the SEC for the Ben Graham: Net-Net Newsletter. This is probably a mistake. But it’s a mistake I think readers understand. Most of them have never looked at companies that don’t file with the SEC.

If you look at this particular list – you’ll see that there were only three companies with a market cap over $10 million in April 2009. And only one company with a market cap over $100 million.

Overall, if you generated a list like this, did your best to buy shares, and then held the stocks for two years without ever looking at them in the interim – you’d do very well.

How many people would actually try this?

Almost nobody.

The same list – of the least well known net-nets – for today has only five stocks with a market cap over $10 million. And one of these is a Chine reverse merger stock. And this one – I’m not going to say the name – is clearly a total fraud. (Hint: You can tell it’s a fraud by looking at their inventories and receivables relative to sales and current liabilities. Real companies don’t look like that. A company about 1% to 2% the size the company claims to be would look a lot like that).

But it also includes two of the stocks in the Ben Graham: Net-Net Newsletter’s model portfolio. The list I showed you from two years ago also has two stocks now in the Ben Graham: Net-Net Newsletter’s model portfolio.

So, although I don’t run this particular screen to find net-nets for the newsletter – I’ve ended up with one-third of the newsletter’s portfolio in stocks that passed this screen (of the 13 most neglected net-nets) either in 2012 or 2010.

If I ran the screen for 2011, I’d find yet another stock in the Ben Graham Net-Net Newsletter’s portfolio on the list.

Last year’s list of most neglected net-nets also provides a good warning. Four of the 13 stocks are up for the year. None is up less than 90%. One is up 400%. So, there are your big winners.

But 9 of the 13 most neglected net-nets from last April are down for the year. Those are your losers. And five of the nine losers are down less than 10%.

That’s the key to net-net investing. Over a single year, something will happen to a couple of your net-nets to give you huge gains. Nothing good will happen to the rest. Some will fall lower in price. Some will fail.

It’s less important to look at how many rose or fell over a year. It’s more important to look at how many net-nets lost almost all of their value. And how many rose faster than their NCAV and no longer qualify as net-nets.

Backtests support the idea you have two objectives in net-net investing:

1. Trying to own big winners when they have their moment of glory

2. Trying not to own big losers when they fail catastrophically

On point No. 1, selling your net-net within one year of buying it is often going to result in you watching as that company gets bought out in a year or two or three. When you no longer own it. One reason why you have big winners in net-nets is that some specific event happens within one year like a takeover. But the stock doesn’t know you own it. Half of all net-nets will be net-nets again next year. There’s no reason the wonderful event that gets you your return needs to happen within the first year of you owning the stock.

For the sake of argument, assume that there’s a 50% chance a net-net will still be a net-net next year. If you buy a stock when it becomes a net-net your chance of it still being a net-net in future years might look something like this:

· 1 Year Later: 50%

· 2 Years Later: 25%

· 3 Years Later: 12.5%

· 4 Years Later: 6.25%

· 5 Years Later: 3.13%

That’s just an illustration. It’s taken from the backtested knowledge that a basket of all net-nets only has about 50% annual turnover. So half of the stocks in a net-net basket tend to reappear 12 months later on the same list of net-nets. The important point from my perspective is that you need to wait until years three or four to have a portfolio where the vast majority of your net-nets have either thrived or died. If you sell your net-nets after one year, you’ll just be cutting down half your crop of net-nets without ever seeing whether they had something good or bad happen to them. They’ll simply still be net-nets.

For a net-net, to be a true success or a true failure we know one thing – it can’t still be a net-net. It either has something happen to raise its price above NCAV. Or it becomes a total loss. It fails financially. In either case, I can promise you there will come a time when it’s not a net-net.

So, to me, it looks like a net-net portfolio with 100% turnover is simply chucking half of its developing net-nets into the trash each year for no good reason. Even a net-net portfolio with a holding period of two years is probably discarding one out of every four developing net-net stories for no reason other than an urge to be active.

I think inactivity is necessary to really judge your net-net selection skills. Because we know net-nets either result in some very big good thing or bad thing happening to the company. Like the company gets a buyout offer, turns its business around (and is valued on a P/E basis), loses that key customer or falls into bankruptcy.

When I look at the numbers, I think they point to a three to five-year timetable. Even then, not every net-net you pick will have ended in a clear success or failure of some sort. But most of them will. At one and two years, I really do think we are talking about 25% to 50% of the net-nets you sell being incomplete trades on your part. You bought the stock, but then you didn’t stick around to see what happens. And to me, if you bought something when it was a net-net and you sold it when it was a net-net that means you didn’t let the situation you originally bet on play itself out.

Now, if you’ve found a better opportunity somewhere else, or you need the cash, or something like that – I understand and agree with the sell decision. But if you are just taking a profit (or loss) in a net-net within 1 to 2 years of buying it – I think you’ve basically just traded a stock on a different basis than you analyzed it.

You probably analyzed the net-net in terms of the risk of it going under and the reward it could provide if it was sold to a control buyer.

But neither of those things happened if the stock is still a net-net.

So I think a lot of what we see in net-net backtests is really only half the story. It’s showing you what happens to the net-nets that succeed or fail. But because these net-net backtests are often using a one year holding period – we’re often looking at something where basically 50% of the net-nets just did nothing. We just traded in and out of them for no reason. We didn’t stay for Act 3.

I’d prefer to look at three to five-year backtests. The problem with that right now is that we’ve had a really unusual three to five years in the stock market. So, a three-year backtest from today would make net-nets look too good. And a five-year backtest would make them look absolutely lethal to your portfolio.

The one thing I really like about doing a three to five year backtest of net-nets is sorting them by their return over that period. Unlike a lot of backtests of other strategies where you have a real mix of multi-year performance – a backtest of net-nets held for three years or more clearly separates the winners from the losers. You get to see actual resolutions. There will be a few bankruptcies, a few buyouts, etc.

I also want to say something about what backtests can tell you and what they can’t. I get a lot of emails asking me whether I’ve tested something and what the results are.

Really, the results of a backtest should just be a proof of concept. And a good strategy should not depend on any specific way of expressing that concept in formula form. It should work even if you tweak the formula a bunch of different ways that don’t change the concept you are expressing.

I’ll give you an example of something I have tested. And it’s an odd thing to test.

I’ve tested what would happen if you bought net-nets with the highest combination of insider ownership and age (as a public company).

The difference between the oldest and most insider-controlled companies and the youngest and least insider-controlled companies among net-nets is dramatic. The young and not insider-controlled companies still beat the market. But they do it with huge winners, huge losers, as many down years as up years, and a spotty year by year record against the S&P 500. The oldest and most insider controlled companies have far, far fewer big losers. In fact, the average worst loss in a portfolio of 10 old and insider-controlled net-nets is actually less than a more mainstream strategy like Joel Greenblatt ’s magic formula (which actually tends to have big losers). The winners are still big. For every dozen old and insider-controlled net-nets you buy, one will go up around 100% to 300% in the next 12 months. If we go through the record we find that’s usually a buyout.

My point is definitely not that you should keep an Excel sheet with the age of every net-net meticulously recorded in one column, the amount of insider ownership in the other column, and try creating a magic “old and insider controlled” net-net formula of your own.

My point is just that this silly little ranking system proves a bigger concept. That bigger concept is that a good net-net is really just a decent business. Why would insiders still control 10% to 50% of a company that has been public for 10 to 30 years?

Because it’s a decent business. It may not be a great business. I’ve seen insiders happy to hold something earning 6% to 8% on equity for decades. And they think their lives are pretty good. And they’re probably not wrong. I’m not sure you want to get stuck in that stock forever. But a basket of stocks like that is probably not going to end badly.

So, it’s really not just about market cap. If we look at a list of the oldest and most insider controlled net-nets available today, they are mostly above $10 million in market cap. Only one of the top 13 is over $100 million in market cap. They tend to be in the $20 million to $70 million market cap range at the moment. They’re generally liquid enough for an individual investor to own without trouble. Although they aren’t liquid enough to bid for in the morning and know you’ll get your shares by 4 p.m. today.

These are the kinds of net-nets I buy for the Ben Graham: Net-Net Newsletter. They aren’t always insider controlled. But they’re almost never very young companies. They aren’t always consistently profitable. But they usually have no more than a couple losses in the last decade or so. They aren’t always overcapitalized. But they usually have more cash than they need.

Each of these “rules” has been broken by at least one of my picks for the Ben Graham: Net-Net Newsletter.

Two of the stocks we bought have nothing but a series of small (relative to their cash hoard) losses because they have cash that was not earned by their operating business. We bought those for the cash.

A couple of the stocks we bought have almost no insider ownership. They were bought at good prices relative to cash, receivables, etc., (not just inventory) and have a long history of at least eking out minor profits. Right now, these are the stocks I’m most worried about. The lack of insider ownership worries me. And these companies have made constant capital allocation errors. Ironically, the names of the net-nets in this group are probably the best known and most written about net-nets on other blogs, websites, etc.

And last month’s net-net – which is an old (for its industry) company controlled by the same CEO for decades – violates the rule about leverage. It’s got a lot of bank debt. And is maybe the riskiest net-net I’ve ever picked for the newsletter. It’s also very cheap.

I can’t backtest a formula the newsletter applies – because I don’t use a formula. The best way to explain it is that I try to buy the “best” net-net available each month with an emphasis on “safe” rather than “cheap.” What does best mean? It means attractive to me. There’s no technical definition.

Some of the things I find attractive in a net-net are:

· Long history of consistent profitability

· Long tenured CEO

· High insider ownership

· Low leverage

· High cash relative to share price

· Good capital allocation decisions

· Simple business

· Good business

· Lasting business

Ideally, I want a company that does the same thing year after year – something it’s been doing for a few decades – and earning money doing that. Ideally, I don’t want to pay more than the company’s own cash for the stock. We almost always do. There’s really only one stock in the portfolio now that is a decent business we bought for just its cash. Everything else we either bought for cash (but didn’t get a decent business as part of the deal) or it’s a decent business but we had to pay more than the company’s cash per share.

I’m not sure how to backtest that approach. Depending on which part of that recipe you focus on, you’ll get different results. High insider ownership alone doesn’t tell you much – Chinese reverse mergers sometimes have high insider ownership. But those are young companies where the insiders haven’t had a decade or two to sell their shares. High insider ownership in a newly public company doesn’t have the same meaning as it does at an old company.

If you just ranked net-nets by lowest leverage, all you’d get is cash shells like Cadus (KDUS). That’s fine. I’m not saying it’s a bad net-net. I’m saying it’s an atypical example of what we buy throughout the year. You’d get the same result if you looked for cash per share.

I don’t know how to test good capital allocation decision. If you did a screen for net-nets paying dividends – yeah, you’d find some of what’s in our portfolio fairly fast. I’m a sucker for net-nets that pay dividends. Not for the dividends themselves. But for everything those dividends are a symptom of.

Lasting business is impossible to test for. It’s absolutely critical. But how do you separate technologies that are here to stay from those that will be gone in five years.

There is one point to make on this issue though. I hate retail net-nets – even though I think they usually provide the biggest upside. We avoid those in the newsletter. The reason for this is the “lasting business” issue. A retailer that loses some of its competitiveness is a retailer that could be out of business in a matter of years.

Good business is something you can test for. You can use something like Joel Greenblatt ’s definition of return on capital. We have two net-nets in the portfolio right now that would definitely be near the top of the list in terms of having very high ROCs for a net-net. However, I look at 10-year averages. Not just one year.

I think the idea of backtesting this approach is not the issue for would be net-net investors. The approach works. I know that not from backtests of this specific approach but from a lot of evidence – including backtests – that shows almost any approach to buying net-nets will work. It’s actually hard to find ways to formulaically sabotage a net-net portfolio so it underperforms the market.

But it’s very easy to sabotage an actual net-net portfolio run by an actual human being.

All you have to do is sell too soon. If you trade too much – and especially if you sell your winners when they go up in price, you can destroy the performance of a net-net portfolio.

But you could do the same thing to a Peter Lynch or Phil Fisher approach. If you sell those kinds of stocks when they rise 30% or 50% or 100% – you can very quickly lose money on an overall portfolio of those kinds of stocks.


Because they need big winners to offset the losers.

With net-nets, I think people are focused on the idea that it’s a quick trade. You buy it. It rises. You sell it. You don’t want to get stuck in the stock. It’s a lousy business. It’s a “cigar butt.” And so on.

I keep reminding people that Ben Graham and Walter Schloss had much longer average holding periods than the typical mutual fund has today. We are talking about guys who held a net-net probably 3 times longer than a mutual fund holds a “high quality” stock today.

If you’re not doing that – you’re putting a twist on their strategy. It could be a good twist. Or it could be a bad twist.

But I don’t think the actual net-net selection method is critical. I think the holding of net-nets is critical. I think getting enough out of your winners is critical.

It’s a mistake to think that a backtest that proves a certain net-net strategy works means that same strategy will work for you. It won’t. Most net-net strategies will outperform the market. Most investors who follow them will underperform the market.

They’ll sell too soon.

So, before deciding if following the Ben Graham: Net-Net Newsletter is the right approach for you, you should think more about an emotional backtest. Would you really have held some of these past winners long enough to get enough out of them?

Or would you sell a net-net the newsletter chose to hold on to after a year. Because if you’d sell some net-net even when the newsletter recommends holding it for another year, I really don’t think your results will have any resemblance to any backtest I can run.

Backtests don’t take into account the human tendency to “cheat on their diet” so to speak. But most people are going to do that. And their results will reflect it.

As far as whether a focused or non-focused net-net portfolio works best – in backtests a non-focused portfolio almost always works best. The reason is simple: there’s often a 1,000% winner in there somewhere. I’m not sure how realistic that is. For one, it’s usually a tiny company. It’s usually something that just traded down at a crazy price for a little bit. It’s pretty much an anomaly that you could have maybe bought – but would you have held on for the 1,000% increase?

The issue between focused and non-focused net-net portfolios is more mathematical than practical. When you have some insanely huge gains in a population of stocks, picking one of those will often improve the portfolio, and a less focused portfolio allows you to do that (by pure chance). At least in a backtest.

I actually don’t think it matters how many net-nets you keep in your portfolio. I think it matters how fast you add and subtract them. My advice is to limit how many new net-nets you buy. It can be one a month or one a quarter or something like that. That’s fine.

Don’t try to buy 10 net-nets all at once, unless you are actually using a totally mechanical strategy you know you will stick to. Very few people can do that. Backtests show that it’ll work fine if you buy 30 net-nets all at once without regard to their quality.

But would you really trust that portfolio enough to hang on to it in a period of underperformance?

I think the best net-net portfolio is the one that is easiest on you psychologically. For most people, that means “collecting” net-nets at a rate of about one a month to one a quarter and then not worrying about revisiting them until the first couple years are over.

That’s what I’d suggest to someone getting started in net-nets.

Slowly assemble a net-net portfolio over a period of a couple years.

And don’t sell anything for those first two years.

Read Geoff’s Other Articles

Ask Geoff a Question

Check out the Buffett/Munger: Bargain Newsletter

Check out the Ben Graham: Net-Net Newsletter
Not a Premium Member of GuruFocus? Sign up for a free 7-day trial here.
4 / 5 (12 votes)

Please Login to leave a comment