The never-sell stocks list

Over the last 20-odd years we've recommended companies that have gone bust (Timbercorp), changed our minds on former buy recommendations and locked-in a loss (Unibail), been late to spot under-investment in product (GBST), fallen in love with management (Roc Oil) and got our numbers wrong (Timbercorp again).

Although it rarely occurred, in each case the worst possible outcome was a total wipeout. What could possibly be worse?

Loss aversion aside, how about Cochlear, a five-bagger over the four years from when we initially recommended it on 3 Jul 98 (Buy - $6.30)? Locking in profits in 2002 seemed sensible, especially with an apparently high valuation. And anyway, who could possibly be dissatisfied with a 550% gain?

Every Intelligent Investor analyst, apparently. Having asked them to nominate our greatest investing mistake over the past few decades, selling out of great businesses got number one spot, with Cochlear the prime example.

Had we managed to set aside our concerns about valuation, that five-bagger would have become a 32-bagger. You can add up all the losses from the likes of Timbercorp, Unibail, Roc Oil and GBST etc. and all would have been overcome by this one fantastic gain. This was the key lesson from six years ago when Research Director Nathan Bell confessed his worst calls ever.

Great investment track records tend to be determined by a few great ideas. Murray Stahl analysed Peter Lynch's famous track record of earning 29.2% p.a. between 1977 and 1990, with a view to understanding how a portfolio containing 1,400 stocks could do so well. Surely a thousand plus stocks would be a proxy for market returns?

Instead, Stahl found that just two stocks, both of which emerged from bankruptcy, accounted for most of the performance. That's two stocks in 1,400.

Our track record, good as it is, would be more impressive had we hung on to Cochlear. In our defence, we have jumped back in a number of times over the last 15 years, but at the expense of the many advantages of holding great stocks for the very long term.

These extend well beyond the old saying that good things tend to happen to good businesses.

As Graham Witcomb wrote in Why the dead outperform the living, "A University of California study of 66,000 investors found that the higher the portfolio's turnover, the lower the average return. Those who traded the most lagged the overall market's performance by 6.5%. As the researchers put it, 'trading is hazardous to your wealth".

Hanging on to good businesses lowers transaction costs and allows a growing company to compound your investment over time. Cochlear is a great example.

The second advantage is covered in James Greenhalgh's Till death do us apart - never sell list (Editor's note: my best story ideas are always stolen), who wrote that:

"Selling apparently highly priced but excellent businesses has usually turned out to be a long-term mistake. Wonderful companies can end up creating value over time in surprising ways, whether through internal investment or acquisition. Often it's simple mathematics. Companies that can reinvest capital at high rates of return can compound value significantly over time".

Selling out usually triggers a tax liability that reduces the impact of compounding.

The third is via Nathan, who makes the point that "it's difficult and time consuming to keep finding new opportunities to earn high returns after paying tax on your gains from previous ideas".

Selling companies you know inside out to buy companies you're less familiar with is risky because of what Nathan calls 'shiny new toy' bias. "People overestimate the value of new ideas over old ones simply because they're more mentally stimulated by new ideas."

That's something we were perhaps guilty of in the case of Cochlear. High growth over many years is more valuable than most people realise; selling such opportunities on concerns over valuation alone is usually a mistake.

Whilst Cochlear was our clearest mistake in this regard, we have done better elsewhere. Having recommended CSL on 20 Jan 10 (Buy - $31.30) we've held on up to $230. As Graham Witcomb notes in the most recent review, "CSL can't be valued on its current earnings alone - its competitive advantages and long, steady runway for growth are where most of the value lies."

The lesson, in the words of Gaurav Sodhi, "is to think about business strategy and competitive position more than you think about statistical cheapness." Or apparent expense.

Adopting that mindset has allowed us to make substantial gains on high quality businesses like CSL, ResMed, ASX and Sydney Airport. Our track record on Sonic Healthcare, Macquarie Group, Cochlear and ARB, where one small, caveated Sell recommendation was arguably a blemish, is less distinct.

Almost every company faces structural challenges. It would therefore be foolish to say there are stocks that one should never, ever sell. But it makes sense to at least have the intention to own high quality companies with deep competitive moats forever. This isn't to disregard valuation, but to weight it appropriately, in the manner of Graham's comment on CSL.

It is also to make the point that a few high performing stocks will likely account for most of the returns in your portfolio. Those stocks are more likely to be high quality businesses that can compound returns at impressive rates over long periods. The eternal problem is that we can't possibly know exactly which stocks will be the high performers.

If a few good choices will make most of your money and you can't pinpoint them it makes sense to at least have a mental list of those companies most likely to jump the high quality bar. With that in mind, our never-sell stocks are listed in Table 1.

To access more share research from Intelligent Investor, start a 15-day free trial.