I originally wrote on AAON Inc. (AAON) in June, arguing that it was a great company, but overpriced at the time. 6 months, 2 great quarters, and a 40% increase in stock price later, I agree more than ever with both of my original assessments.
A Great Company- Don't Forget it!
Before I move on, I want to emphasize that first point again- I absolutely do believe AAON is a great company. The company makes great products. My university has AAON HVAC units all over the place and they work wonderfully to the best of my knowledge in Buffalo, an area prone to high winds, snow, rain, rapid temperature and humidity changes, and just all-around crazy weather in every sense of the words. Just the fact that AAON can secure the business of a major state university with 30,000 students is a testament to their products, reputation, and reach. These past few quarters, the company has been able to report huge increases in all the right areas of the income statement- much a result of increases in market share in a very competitive HVAC market filled with much larger companies than AAON that have access to the capital of much larger companies that they are subsidiaries of.
The remainder of this article is therefore not intended to take anything away from AAON the company, but to point out that AAON stock is way too expensive at this point. Management doesn't have much control over that and should not be blamed. Also keep in mind- stocks rise a fall. AAON stock is very expensive now but there's a very good chance that it will be attractively priced sometime in the next five years, especially because of the clear cyclicality of the business. Keep this company in mind and take a look at it once in a while so you're prepared and able to capitalize on such an opportunity.
The rest of this article will argue that there is a major price inefficiency in AAON stock right now- much greater than when I first analyzed the company 6 months ago.
I recently read Valuation by McKinsey & Company. The book is extremely comprehensive, covering not just the process of modeling and valuation, but also valuation in special situations- like valuation of a cyclical company. It also provides McKinsey studies of historical market behavior towards such special situations. The McKinsey consultants found that markets tend to almost completely forget that cyclical companies are cyclical. As the impressive, improving numbers come in near the peak of the business cycle, investors (and sell-side analysts even more so) extrapolate the great results into the future, way past when they should know things will get tough again, creating unjustifiably high valuations that don't last once industry pricing and demand weakens and the financial results revert to the mean. I don't want to paint a picture of everyone doing it, because everyone doesn't, but most money does and in a short-term voting machine of a sentiment-driven market where everyone has some say, it only takes a little bit of ignorance to have a big impact on prices.
I think that's what's happening with AAON right now. It seems that the market is in awe of the numbers that AAON is putting up right now and assuming that the train will just keep rolling. While the company is certainly performing strongly and taking market share on an absolute basis, you'll note that in each of the last 4 quarterly earnings calls, price increases have been attributed by management as a major factor in the growth. Management is blatantly telling investors cyclicality is involved, so the blame can definitely not be put on them for intentionally 'talking up' their stock and taking credit for an improving business environment. The blame is on investors for ignoring these warnings and justifying their investments at this price level with assumptions about the future that just won't come to be.
If you look at the critical metrics, you'll see that, yes, relative to past cycles, things are improving faster and will likely peak somewhat higher than at the last peak in 2009, but we're still well into the cycle, past the optimal time to buy, and nearing when regression would typically occur.
From the cycle it looks like the regression is about 2 years off, and yet as I'll point out later, to justify today's price in a model I have to extrapolate the current trend and pretend cyclicality doesn't exist.
Fiscal 2012 was very strong for the company, and 2013 started off the same way, but the company has seen sales trends do a complete 180 and expects Q4 profits to be the same as Q4 2012 on lower sales.
Backlog is the best leading indicator we have for a company like AAON, and the data from the last 3 quarters tells me that the sales environment is getting much weaker very quickly. This data suggests that the business is entering a weak period again, possibly sooner than in the 2 years that the chart would indicate. Whether the downtrend is in 2 years or less than 2 years, this does not bode well for the next couple years.
I went about valuing AAON stock by several methods and ended up concluding that the stock was very expensive.
For multiple analysis, I much prefer EV/FCF to P/E, EV/EBITDA, or any other metric. I feel that enterprise value is the truest measure of market value for a non-financial firm and allows for benchmarking independent of capital structure. In the same way, I think free cash flow is the truest measure of profits for a non-financial firm.
Just on a trailing basis, AAON has an EV/FCF of 24, which looks rather expensive, but not outrageous. However, again this is an obviously cyclical company and so I wanted to capture that cyclicality. When valuing a cyclical, erratic company the idea is to capture typical or average performance. Right now the company is much closer to a peak than a trough in the progression of its business cycle, so the trailing numbers are not and should not be though of as typical. What I didn't want to do though is just use a 5 year average FCF because AAON is also a fast growing company, and pulling in data from 5 years back would understate the company's current profitability. Instead, I computed the company's average FCF/Revenue margin, 8.44%, and applied that to trailing revenues to come up with a normalized FCF number that is more independent of the business cycle. The company's EV/Normalized FCF multiple of 38 is much more telling. Very few companies (none that I've encountered) deserve such a high multiple and AAON is certainly not one of them. It's very difficult to determine a specific multiple that the company 'deserves' but if you put a gun to my head, I'd say something like 19-21 would be appropriate based on growth, profitability, and the overall quality of the company and its products. Based on that, the EV/Normalized FCF multiple indicates around 100% overvaluation.
I also valued the company using a dividend discount model. Right off the bat, AAON presents a challenge to this method because the company pays a dividend but also returns a good deal of cash to shareholders via share repurchases. More and more companies are choosing this alternative to paying dividends in order to capitalize on extended periods when their stock is believed to be undervalued, in order to defer capital gains taxes for shareholders, and give shareholders more choice in the matter (they can effectively create their own dividend by selling 2-3% of their shares during the year at opportune times when the stock is most expensive). Valuation expert Aswath Damodaran, whose models I use in my own valuations, believes the best answer to this problem is to assume a share repurchase is a dividend. It's not quite as simple as that though. Not all share repurchases are created equal. When a company finances a buyback with net income, that fits the bill as a dividend equivalent- it's a return of capital, but when the company finances the buyback by issuing debt, that is not a dividend equivalent or return of capital because the company did not earn the money it rewarded shareholders with. The solution is to reach back at least 5 years and figure out what net buybacks (buybacks less debt issues) and dividends have been relative to net income in order to determine a normalized payout ratio to input into the model. Doing this for AAON, we find that much of the buybacks have indeed been financed by earnings, resulting in a much higher adjusted payout ratio (69.1% vs. 26.3%) and in turn a more accurate, conservative valuation.
So there you have it, a lot of words, but that's my first assumption:
- Normalized payout ratio of 69.1%
And here's the rest:
- Normalized ROE: 24%
- Risk-free rate: 2.88%
- Equity risk premium: 5.5%
- Current beta: 1.21
- Terminal beta: 1.2
- Length of high growth period: 10 years
- Terminal ROE: 10%
- Gradually adjust ROE and payout in year 6-10 to terminal values
- Terminal growth rate: 1.5%
The resulting value was $12.35 per share, 58% below market. Because of the huge discrepancy, I went back and changed the payout to the original 26.32%, and made some other bullish, 'throw caution to the wind' assumptions:
- Normalized payout ratio: 26.32%
- Length of high growth period: 15 years
- Terminal ROE: 15%
- Terminal growth rate: 3%
Even with all this, I still only came up with $28.37 per share, or about 4% less than market. I tried to play the extreme bull 'no such thing as cyclicality or share repurchases' game and still could not completely justify the current market price.
On to book value. The company currently sports a book value of $162mm or $4.41 per share. The stock's price/book is 6.98. Combine that with the normalized ROE of 24% that I assumed in my model, and the effective yield is 3.44%. That's the return investors can reasonably expect buying AAON stock at this level and holding it long-term. That's barely better than the risk-free rate and loaded with all the risks and uncertainties associated with an equity. The company's beta of 1.21 and an equity risk premium of 5.5% suggest investors should accept nothing less than an effective yield of 9.54%. To get that investing in AAON, the stock would have to trade at 2.52x book value, or about a third of where it is right now.
If you look at that spreadsheet screenshot I posted above again, you'll see that while the company has repurchased shares in the first 3 quarters of fiscal 2013, the extent of the repurchases is much lower than what the company has historically done. Taking the prior 5 years together, multiplying by ¾, and determining an average share repurchase dollar amount for 3 quarters, I get $8401k, or about 74% more than what the company has devoted to buybacks in 2013 thus far. I think this, along with some recent insider selling supports the idea that AAON shares are very expensive right now. These are important things to look at in my opinion to get a sense of what a company's management team thinks of share price relative to underlying value, but I also don't want to look too much into this. First off, insiders can sell shares for a number of reasons, and the selling hasn't been so extreme that I'd consider it out of the ordinary. Second, the repurchases have been trending lower in recent years anyway and the company seems to have other plans for its excess capital. An analyst brought the share repurchase topic up to Norman Asbjornson during the Q3 call, to which he replied:
We aren't talking special dividend. What we are doing is we're doing a couple things. Number one, we're completing some of our capital expenditures in the building program. We've got a long-term plan. We've laid out in great depth several years in advance where this company should be going. And so we're going -- moving forward with some of those building plans as we speak. Probably the biggest one that we have coming up, and we are not ready for it yet because we have to do our homework, so to speak, to make sure we've studied it out well enough to do it correctly, and that is we need a new laboratory. And the new laboratory is going to cost an unknown amount at this time, but it's probably going to be in the $10 million to $15 million arena or somewhere in that vicinity. And that could occur next year. So we need to keep that around. We're also thinking that contrary to what we normally do, there might be an opportunity to buy something, and it doesn't hurt to hold money. And the third thing, which is very obvious, is we're talking about should we increase our dividend. And so all those things are on our plate, and we're going to be trying to come to a conclusion in the next quarter as to what should we do, particularly as it relates to the dividend.
Sidoti & Company's research reports on AAON are posted on AAON's IR site. According to the report, Sidoti rates every company in its universe as either a buy or neutral. Of all the companies it rates, 66% are buys and 34% are neutral. Despite being seemingly very generous with buy recommendations, Sidoti has a neutral rating on AAON stock and a $24 price target. The analyst believes AAON is 'fully-valued' and that the recent positive results 'are already reflected in the price.'
To conclude, I still consider AAON to be a great company, but now feel that the stock is quite mispriced. Historical data and a recent deterioration in backlog suggest that the company is dealing with a weakening sales environment that will get tougher in the next few years as part of a regular and predictable business cycle. Price inefficiencies on highly-cyclical companies are all too common, and that seems to be the case here. The current rich valuation can only (and barely) be justified by assuming that the business is no longer cyclical and that the current fluctuation upward in profitability is not a fluctuation but a sustainable trend that will continue for years to come. I think shorting would not be such a bad idea at this point but at the very least, I would not go long.