Someone emailed me this question:
"I found your use of a multiple to estimate the value of NIC (EGOV, Financial) in five years interesting (I often do the same thing by using the company’s historical normalized multiple). The logic is sound, but in a way, isn’t that sort of blending a trading philosophy in with an investing philosophy?"
Yes. It is like blending a trading philosophy with an investing philosophy. The theoretically sound way to make an investment is to use a discounted cash flow analysis (DCF). However, a DCF is not practical. I talk to a lot of investors and even when they tell me their approach is probably "too simple," it really looks too complicated to me.
I do put past historical data into Excel to see it all in one place and do long-term calculations to create a "Value Line" type data sheet. I also have a calculator with me at all times to let me quickly do calculations I might otherwise not bother doing. If your choice of which stock depends on something Excel calculates for you (as it will in the case of a DCF), it's not simple enough. Honestly, if it's something you need a calculator for, it's not the right investment. The "proof" of which stock to buy should be something you can do in your head.
Let's take the NIC example. I'll make up these numbers so I don't need to worry about finding the exact figures for this example, but it's close enough to illustrate the point and relate to NIC.
Let's say NIC now has a 4% free cash flow yield. However, it is diluting its share count by something like 1% a year in a normal year. That's 4% minus 1% equals 3%. If NIC uses some of its free cash flow to offset all of that dilution, that leaves us with 3% of our possible purchase price (as of today) being paid out in a dividend or used to reduce the share count.
Now, let's say NIC will grow total revenue (so same state revenue plus new federal contracts, additional states, etc.) by somewhere between 8% and 10% a year for the next five years. That would mean the stock would return 11% to 13% per year over the next five years. That's only true if the multiple doesn't contract after that point.
Let's think about the "terminal growth rate" we'd expect in NIC. Inflation is probably around 3% normally and nominal gross domestic product (GDP) in the U.S. is probably around 5% (inflation of 3%, population growth a little under 1%, real output growth per person a little over 1%). NIC will try to win some new business. It might be able to do a nominal GDP type growth rate. It might only be able to do an inflation type growth rate, but it probably won't fall too far outside those figures. Let's use 3% to 5% per year as the rate NIC will grow in perpetuity after the last phase of its fast growth period.
What should a stock trade at that grows 3% to 5% a year. Well, NIC can probably grow 3% to 5% a year forever without retaining any earnings. It quadrupled sales over a decade without really increasing net tangible assets. So, all earnings are potentially able to be paid out in dividends. That means the stock's "fair value" multiple in the sense of pricing it to perform in line with the Standard & Poor's 500 would be: (Expected Forward Total Return in the S&P 500Â –Â NIC's Perpetual Growth Rate) gives you the earnings yield at which the stock should trade. However, it does dilute by about 1%. So, we have to knock 1% off this number. Basically it's S&P 500 Expected Forward Return minus 2% to 4% equals the earnings yield at which it should normally trade in the future. The average return in the S&P 500 from 1928-2016 was about 9%. Let's use that. I expect much lower in the future. This is conservative.
9% minus 2% equals 7%.
And 9% minus 4% equals 5%.
1/7 = 14 (roughly)
And 1/5 = 20.
The stock should trade at a price-earnings (P/E) of 14 to 20 once it is growing as slow as 2% to 4% per year in terms of earnings per share growth. That's because a stock growing EPS by 2% to 4% a year and paying a dividend yield of 5% to 7% per year will return somewhere between 7% and 11% per year. That should match the market. Actually, it'll tend to beat it.
What's important here is how simple and conservative this approach is. That's why it's much more useful than a DCF. Beyond five years, let's see what we've assumed:
* The company will grow overall sales, earnings, etc., at a rate no less than inflation but no more than nominal GDP.
* The company will pay out 100% of its earnings in dividends or share buybacks.
* The S&P 500 will return somewhere between 7% and 11% per year –Â and we are using that as our opportunity cost.
* The company will always dilute its shares at 1% a year forever.
None of these estimates is aggressive. Now, that leaves us with the issue of how well NIC stock will perform over the next five years if it slows down to a growth rate of just 3% to 5% per year before dilution and 2% to 4% per year (per share) after dilution in five years.
The company is guiding for EPS of 69 cents to 72 cents next year. Since this is lower than its most recent year of EPS – we'll use that. Let's apply a 7% to 9% (8% to 10% less 1% of dilution) growth rate for five years to that figure.
69 cents * 1.07^5 = 97 cents per share
72 cents * 1.09^5 = $1.11 per share
I said a long-term normal multiple of 14 to 20 would make sense for a company growing EPS as slowly as 2% to 4% per year forever but paying out all earnings in dividends.
That's a price range for 2022 on the stock of:
97 cents * 14 = $13.58
$1.11 * 20 = $22.20
The stock is now at $21.05 per share. The stock is overpriced if we expect EPS growth to be as low as 2% to 4% per year starting as soon as five years from now.
What I just laid out here is very conservative. It's possible NIC could really do something wrong and lose some contracts or something. Short of that, the stock trading at lower than a P/E of 14 to 20 in sooner than five years is really unlikely.
What we did here is really very similar to a DCF. We broke the company's future into two parts we think we can understand differently. One, the next five years where we think we know it will grow faster than many businesses do. And then – beyond that – we have it growing quite slowly. Now, if we made a mistake and said that growth would slow to 2% to 4% per share in just five years when in reality it will continue to grow at 7% to 9% a year for more like 15 years, qe have badly miscalculated. Also, if we said per share growth in EPS would be something like 2% to 4% a year forever and it turns out to be 5% to 7% a year forever – we've badly miscalculated.
In fact, by my estimates, if NIC could grow free cash flow per share by 5% to 7% a year on a long-term basis while also paying out all earnings in dividends, it would almost certainly outperform the S&P 500 even if priced as high as 25 times earnings. That's because the S&P 500 will do worse than 9% a year.
I think it's very fair to say what we've done here is estimate what a five-year trade would look like, but I think that's realistically as far as we can make any sort of practical estimate here.
I've written before that you should make "point-to-point" estimates of what a stock will return over anywhere from 5 to 15 years. I don't believe in using calculations that aren't static (in the sense they involve a definite end point). Calculations that go out less than five years or more than 15 years aren't helpful.
Why aren't calculations of less than five years helpful?
Because value will beat quality, growth, etc. over periods between now and five years from now. It's just not worth doing the calculation. You should just buy whatever has the lowest valuation relative to intrinsic value. Let's say I have two stocks: Hunter Douglas (HDG, Financial) and Omnicom (OMC, Financial). When we picked these for "Singular Diligence," we appraised Omnicom as selling for 90 cents on the intrinsic value dollar, and we had Hunter selling at 47 cents on the intrinsic value dollar.
What if Omnicom and Hunter Douglas reach intrinsic value 12 months after you buy them?
Annual return in Omnicom is 11% from the one-time value adjustment. Annual return in Hunter is 113% from the one-time value adjustment.
Reach intrinsic value in two years.
Annual return boost in Omnicom from "value" is 5%. Annual return boost in Hunter from "value" is 46%.
Reach intrinsic value in three years.
Annual return boost in Omnicom from "value" is 4%. Annual return boost in Hunter from "value" is 29%.
Let's skip ahead to five years.
Annual return boost in Omnicom from "value" is 2%. Annual return boost in Hunter from "value" is 16%. In other words, a 90-cent dollar becoming a 100-cent dollar over five years gets you a 2% annual capital gain. However, a roughly 50-cent dollar becoming a 100-cent dollar over five years gets you about a 15% gain. Quality isn’t going to make up a 13% annual return difference. Over five years or less, you should always buy the 50-cent dollar instead of the 90-cent dollar.
There's no way a "quality" or "growth" stock can compete with a value stock if your value stock mean reverts to the correct intrinsic value within a holding period of about five years. Nothing returns more than about 20% per year long term in terms of its actual compounding of intrinsic value. Even Berkshire Hathaway (BRK.B, Financial) has often been around 20% in terms of long-term compounding. Any time your return from a value gap closing is 20% or more per year, you should just buy the value stock.
That's why I like to look at the situation starting about five years out. If we had a really good growth stock, something like NIC in its best days or Apple (AAPL) in its best days or whatever you want to pick – we could imagine a high-quality growth stock that is only slightly undervalued today running neck and neck with a only so-so quality value stock over five years.
A value stock will outperform a quality stock if the value gap closes in one to four years. For example, Hunter paid a dividend, too. If you bought Hunter at about 50 cents on the intrinsic value dollar and expected to sell it in about five years, you'd be looking at around a 20% expected annual return. That's why value investing works so well. It's very hard to find any stock that will grow 20% per year for five years and not have its multiple contract by the end of that five-year period.
Value investing beats quality investing over periods shorter than five years – and it's simpler just to accept this fact and not worry about calculations about the trade-off between value and growth for holding periods of less than five years. Just find the best, safest stocks you can trading at about 50 cents on the dollar and hold them for five years. They'll beat everything else. And there's no higher math needed to know that.
What about periods beyond 15 years? That’s the point where the situation can realistically flip so quality starts always outperforming value from 15 years on.
Let's say Omnicom will grow EPS (this includes share buybacks) by about 7% per year forever while also paying out a dividend yield of 2% per year. That means the annual return will be 9% per year plus the closing of whatever valuation gap there was. Over 15 years, the closing of a valuation gap from 90 cents on the intrinsic value dollar to the full 100 cents will only be 1%. The 15-year annual return potential in Omnicom is 10% a year.
What about Hunter Douglas? Let's say that back – at a much lower price, when I picked this stock for the newsletter – we thought Hunter was trading for 47 cents on the dollar (a 53% discount to intrinsic value), and we expected Hunter Douglas to pay the same roughly 2% dividend that Omnicom does, but we expected EPS growth at Hunter to only be about 4% per year over the next 15 years. This is in the ballpark of what we'd actually expect at Hunter and at Omnicom (long-term EPS growth of 4% and 6%). Now Hunter has a 2% dividend yield plus 4% EPS growth rate which equals a 6% continuing buy-and-hold-forever type return.
Over 15 years, how much can a valuation gap closing from 47 cents on the dollar to the full $1 of intrinsic value get us? It's 5%. Potentially Hunter could return 11% per year over 15 years versus 10% per year for Omnicom.
What if EPS growth is just 1% lower per year? Then they tie. What if Hunter actually only grows EPS by 2% a year? Then Omnicom wins. Beyond that, we can look at 20 years. At 20 years, the valuation gap closing would provide less than 4% boost to Hunter’s annual returns. And that's from a stock that started selling for less than 50 cents on the dollar. It’s hard to find stocks trading for more than a 50% discount to intrinsic value. It’s hard to ever squeeze more than a 4% annual return from a value gap closing over periods of 20 years or more. As a result, pure value investing that involves a holding period of 20 years or more doesn’t make sense. However, value investing that includes quality and growth makes sense even over holding periods that stretch for several decades.
For calculations of less than about five years, the stock with the bigger valuation gap wins. That doesn't mean lower P/E, price-book (P/B), etc. It means the stock is selling for less than what it'll be worth in five years. A "value trade" over five years. That's a long trade, but, yes, technically it’s still a trade. When I picked Hunter Douglas I didn’t say you had to hold it forever. I said you had to hold it for five years. And we picked Hunter before we picked Omnicom. At that moment in time (Hunter was at a much lower price), we liked Hunter better than Omnicom because Hunter was so much cheaper. Omnicom was always the better business. But, I didn’t think Omnicom was the better stock over the next five years. Hunter was so much cheaper that it seemed the better investment over a five-year time horizon.
And then beyond 15 years, quality wins. If you are holding something beyond 15 years, it's the buy and hold return that matters. So, I think you only really need to do math comparing low-growth value versus high-growth quality stocks in the five- to 15-year holding period range. If you flip everything within five years, buy value. And if you hold everything beyond 15 years, buy quality and growth.
My approach then is to focus on five years as the shortest possible holding period and 15 years as the longest possible holding period. Within that period, I want to keep things simple. I'd like to find stocks that look good both if I buy them and sell them five years from now and also look good if I buy them and then hold them for a full 15 years. I try not to think too much about the next one to four years (because everyone else is doing that) and I try not to think about years 16 and beyond (because those are often difficult for me to know, and the stock’s starting price no longer matters much if I know I'll hold the stock for 20, 30 or 40 years).
We’re mere mortals not machines. It’s simpler for a human to “bound" the problem so it is just a question of what happens if I hold this stock for five years and what happens if I hold this stock for 15 years. I can then assume the actual annual return outcome will fall somewhere between those two extremes. This approach keeps it simple. Otherwise, you fall into the trap of doing a DCF that extends from now to judgment day. Or you fall into the opposite trap of explicitly predicting exactly when in the next five years various events will occur. Both of those approaches are kind of crazy. The stock is tradeable. I’m not going to literally hold it forever no matter what. And this isn’t arbitrage. This isn’t an event driven investment. So, I shouldn’t be predicting if I’ll make money in six months or 18 months or 36 months. It’s one thing to predict certain events will happen. It’s much harder to predict not just what will happen but when it will happen. Ignoring the truly short term (less than five years) and the truly long term (more than 15 years) keeps the problem manageable from a math perspective. Over five to 15 years, I can do all the math I need to do in my head.
I like an approach that accepts the reality that you are likely to sell the stock at some point. And stocks that may generate higher returns in the first five years due to a value gap closing do deserve extra consideration as an investment because you can flip them within those five years and get a higher annual return if Mr. Market offers you the opportunity. You should always go into an investment planning to hold it for the long term, but you should sell one stock and buy a better stock whenever you get the opportunity regardless of how long you've been holding the stock you now like less than what you have the chance to buy. For that reason, assuming I will hold whatever I buy for no less than five years and no more than 15 years is a very practical approach that works well for me.
Disclosures: None.
Start a free seven-day trial of Premium Membership to GuruFocus.
