Right Farm, Wrong Cow?

A Brief History Of Market Mania & Exuberance

The new technological age

This is it, not like last time in the late 90’s. This time, the technological age has arrived. Even though some of the same actors are back (Pets.com has been replaced with Chewy Inc.) this time we are more engaged with technology that the internet has to provide us with than ever before. Also, this time companies, like Google, are making money instead of losing it. You might not remember it but Pets.com IPOd at possibly the worst time possible, February 2000, at $11 a share. It was last listed on the 9th of December 2000 at a price of $0.19. A year later it was completely disbanded.

Cisco, a company that survives to this day, had a run up during the tech boom that saw it become the largest company in the world with a 25th of March, 2000 valuation of $500b. This in the days before a $1t (let alone $2t) company could be comprehended. However exciting many of those stories were however, Netscape is probably the most appropriate example of the moment you know things have changed.

Netscape was the first large company to go public before it made a profit. So, even though the company’s revenue had doubled every quarter of 1995, it was still odd to see the firm listed without a profit to back it up. The shares were offered at an IPO price of $14 and on the first day of trading went to $75. The emergence of a company with no profits, high future hopes and a dream to sell investors became known as the ‘Netscape moment’. This was the dawn of a new industry. Nobody in the halcyon days of the late nineties was to know that Netscape as a company would cease to exist by 2003. All they knew was that a new technological revolution had arrived and they all wanted in. Ironically, the browser that saw the death of the Netscape company, Internet Explorer, would see its own market share obliterated by a new competitor before the second decade of the new millennium was out.

This isn’t even the first time we have seen a technology that would change the world and failed to back the winner. By 1908 when the first mass produced car, the Model T Ford, was invented, there were over 500 car companies in the United States of America. Although investing in Ford would have been a decent idea, you still had to contend with the fact that you had 500 other competitors and you didn’t know, at that time, who would win. In fact, you would have missed the fact that by 2020 the biggest car manufacturer would be Tesla. Retrospectively it is relatively easy to pick the winners, at the time, not so much. Funnily enough, William Durant was founding General Motors at that point and would, as of today, have a higher market capitalisation than Ford.

The last great example is airlines. Anyone that saw the difference between flying and catching a ship between destinations just knew this was going to take off (no pun intended). What happened next? Everyone piled in and hundreds of airlines were started. Mostly doomed to failure from rabid competition, high fixed costs, and exogenous events (think 9/11 or COVID-19) they continued to bumble along as best they could. Many, like Qatar, Emirates, Etihad, are backed by their governments and serve to introduce tourism to their countries. In fact, even Qantas was government owned until its privatisation in 1993 and 1995. With the grounding of 95% of the entire world’s airline fleet it is fair to say we don’t yet know who will make it out of this intact. One thing is for sure, Virgin Australia shareholders are a very unhappy bunch. Their bondholders aren’t a lot better off either. Locally, we have seen the failure of Ansett on September the 12th, 2001, the day that changed the world forever. Since then Australia alone has seen the failure of Horizon Airlines, Great Western Airlines, Airlines of South Australia, Emu Airways (love that name), Aboriginal Air Services, Sunshine Express, Big Sky Express, Transair, O'Connor Airlines, Aero-Tropics Air Services, MacAir Airlines, Regional Pacific Airlines, Tasair, Aeropelican, Brindabella Airlines, and Vincent Aviation. Aviation is just another example of an industry that changed our lives for the better but has been extremely difficult to profit from as an investor.

The Houdini effect. You are seeing what they want you to see

When it comes to companies that offer huge upside potential and have made compelling cases to the market as to why they should and will win you can’t go much further than Tesla.

With Tesla in mind I took a look at what the current share price and profits indicates about the company moving forward. And yes, just like everyone else who was burned for not having Tesla in their portfolio, I have compared it to a tech giant and not a fellow automotive manufacturer. After all, Tesla doesn't sell cars, they sell the dream. The real joke here being that a not insignificant amount of irrational exuberance can go a long way so long as you can get people to avoid looking at the numbers.

If I take a look at Alphabet, Google’s parent company (which should be considered a fair comparison), they are currently trading at a mature, profit generating, P/E of 33.55. I’ll round it up to 35 because 1. I’m a nice guy and 2. Tesla sells dreams whereas Google is basically a pile of shit in comparison, apparently.

If we assume that Tesla ‘should’ trade at a Price to Earnings ratio of 35 and currently is produces 50.47213 cents per share (I know that many decimal places is pedantic but I would like to avoid being accused of bias by Tesla fanboys) then Tesla currently trades at a P/E of 970.06. The share price therefore represents just short of one thousand years of profits at the current run rate. But growth I hear you cry and that is totally fair.

Moving back to where we started; we are going to assume that Tesla grows its earnings so that in ten years’ time (2030) it will trade at a P/E of 35 (greater than Google’s 33.55). In order for this calculation to work it means that the share price needs to stay flat for ten years (unlikely) while the earnings side of the equation grows at 39.40% p.a. every year for ten years. That means that in 2030 Tesla would be spitting out $13.26b in profits every year. For reference, Toyota in the last 12 months made $15.8b in profit and is only rated at a P/E of 14.55.

This should be a lesson that when you are growing a company you can spin a bigger story by having no profits and selling the dream but, alas, it is not. Turns out that you can have a decade of stratospheric earnings baked into the price (and try telling me that an economy growing at sub 3% can support a company’s profits growing a 40% a year for a decade isn’t stratospheric) while selling that same story even when your company has a three figure P/E.

NextDC is dragged into this not because they have an especially charismatic CEO, or a highly publicised mission but because the space they operate in is so highly visible right now. To be a provider of services to online and cloud based companies in the middle of a pandemic that has led to many working from home is to put yourself front and centre on the investment stage. A ten year history of impressive growth and the very alluring potential of a generational shift to cloud based enterprises is too good to pass up. The management team at NextDC have been forthright in their plans and ambitions, it is the market that has gotten carried away.

What business is NextDC actually in?

Here is the trap that a lot of people fall into. Not you, you’re pretty smart, but you know; people. They think they are investing in a segment that could not be more different from the one they are actually in. My favourite example of this is always Tesla. They are a car manufacturer. Everyone invests in them thinking they are a technology company. How’s that working out for me? Judging by Tesla’s $US460b market cap, I’d say not well. The same thing is happening with NextDC.

Many people have jumped into NextDC because of the rush to online applications and the new, Zoom/Teams/Slack way of doing business. In the sense that NextDC services these industries they are right. In the sense that NextDC will have their margins, they are dead wrong. NextDC, as I’m sure you know, offers the physical data centres to these and many other companies and government departments. They have done a great job of centralising the service that would normally be spread between every IT department individually. NextDC, make no mistake about it, runs a very good business. The shame of it is, it’s a steady as she goes business that doesn’t have 50%+ margins.

You see, when you run a technology company like Google or Microsoft, you only put in the effort to write the code once and then you sell that code, be it the Chrome browser or Microsoft Office, multiple times. Especially with many of us buying software on the internet now the marginal cost of each software sale is next to nothing. The difference between that and NextDC is night and day. For every new customer that NextDC adds to its list, the company must provide the equipment, maintain it, run it, and replace it. The company wants to use more, it pays more. It wants to use less, it pays less. There are not the opportunities to make near ‘free’ money as there are in software based companies.

A rate of return measure

First of all, NextDC does a really good job of turning operating cash flow receipts into net of expenses operating cash flow. Before I continue, this might be a good time to look into why I’m using the cash flow statement as a measure instead of the more traditional income statement and earnings per share. Put simply, NextDC has a negative earnings per share number. Normally this would be a bad thing but in this day and age it is actually a huge helping hand as I demonstrated earlier with the Tesla example. It is much easier to tell a story of growth when you are nowhere near having to cash the cheques to investors.

More to the point, you might wonder why I’m using the operating cash flows instead of total cash flows. The problem is that the three categories of cash flow serve very unique and different functions and combining them could lead you to miss something. Let me explain using my favourite example; the fish and chip shop. Operating cash flows are the bread and butter of a business. This accounts for the money spent buying fish and chips and subtracts the costs of running the business. Think sales minus wages, power, and cooking oil. Very simply, operating cash flows are the money you take in minus the direct costs to produce the fish and chips. Investing cash flows are the costs associated with buying new chip fryers minus the cash earned from selling the old ones into the second hand market. Lastly are the financing cash flows. This is where the business accounts for the cash it pays out as a dividend each year to the owners, the cash it brings in by issuing new shares to investors and taking out new loans with banks.

It is now fairly easy to see that by mixing them all together you could get a skewed view of our fish and chip shop. If we are running at a net loss on the operating side (it costs more to produce the fish and chips than what we make selling them) but are selling our fryers and taking out new loans to make up the difference we might think the shop is doing fine. In fact, we would be invested in a business that is only staying alive thanks to equipment sales (and how long can that last) and new debt being taken on. That’s why it is important to at least take a peek at the cash flow statement.

Now that we have covered that I want to get back to congratulating NextDC on how good it is in converting operating receipts into net cash (or operating cash after expenses). It is the sign of a well-run business that NextDC has spent 6 years averaging net operating cash flow around 22% of receipts. This effort on growing the business has culminated in a record operating cash flow receipts of $217m with net earnings of $54m (24.82% conversion). This is a great conversion rate and should be commended. However, and isn’t there always a catch, we need to take a look at how much money has been put into the plant and equipment that generates these receipts to get a fuller picture.

If I look back over the last 7 years NextDC has invested $1.5b in property, plant, and equipment, with $1.1b of that being from before June 30, 2019. The reason I add that last part is this; those predicting ever higher rates of earnings should consider that the bulk of this investment occurred within a time-frame that should have seen it show up in the operating cash flow statement by June 30, 2020. So, based on that $1.5b spent on equipment (the lifeblood of an infrastructure company) the full year 2020 cash receipts of $217m means NextDC earns an operating cash flow yield of 14.1% on its spend. After accounting for the expenses and working on the full year net cash flow of $53m, we can see that the net yield is 3.50% on every dollar invested. If every other part of the business was cash flow neutral and all of this net cash found its way onto the income statement (admittedly, this is a stretch given the income statement is an accounting game) then NextDC would have a P/E of 28.6.

At that P/E, only Spark Infrastructure from the competitors we will list in a moment, has a higher P/E based on figures that try to ‘look through’ the impact of COVID-19. The difference perhaps being that Spark Infrastructure is trading on a trailing dividend yield of 7%, whereas NextDC has no spare capacity to do the same.

Moving back to the expansion of the business I can see that, after the successful closing of a $1.5b debt financing round, NextDC has the equivalent of $1.6b in cash ($893m) and undrawn debt (~$700m). We should pause for a moment and congratulate the management team of NextDC. Given they know they run a highly up front capital intensive business, they have been very clever in their capital raising this year. They have recently lowered their debt costs and extended their loans. They tapped the market for new equity at a time when investors were falling over themselves to invest. This has left them in the enviable position of having a huge amount of capital available to grow even if the capital markets seize up in the near future. If all of this $1.6b in cash were to be ploughed into the business and the net earnings rate increased from 3.5% to 4% (I’m generous) then this would contribute an additional $64m in net operating cash flow. This would bring the total to $118m in net free operating cash flow.

Here is where we separate the well run business of NextDC from the inflated price of their shares. If this more than doubling of free cash flow came to fruition and all profits were paid out to investors, what would that look like? Firstly, we’ll pretend that tax doesn’t exist so that the figures look better for NextDC shareholders. So we have a company that will pay out $118m in cash and is currently worth $5.49b. That is a yield of 2.1%, at best.

Herein lies the problem with investing. Once a share price has had a large run up it is, by definition, a riskier and lower future return investment. A lot of future growth has been incorporated into the current price and, as I just demonstrated, today’s investors need to wait for the full suite of $1.6b in infrastructure investments to be made before they can reap even a 2% dividend. Yes, there are likely thousands of investors who have also seen their investment skyrocket. I am not arguing against that. But if you had invested in NextDC at prices sub $2 then you are sitting pretty with a very large margin of safety. And for what it is worth, that theoretical dividend mentioned above, paid out on shares bought for $1.71 comes to a yield of 14.77%. Of course the people who bought at the lower price have less risk.

Choosing appropriate benchmarking companies

Now that we are on the same page we should take a moment to look around at what some of the other companies in the infrastructure field are doing. The main four we will broadly compare NextDC to are Sydney Airport, Transurban, Spark Infrastructure, and APA Group. Given that Transurban and Sydney airport (toll roads and airports) are clearly impacted by COVID-19, I have elected to use their financial year 2019 annual reports as a standard of comparison. This, hopefully, takes out the impact of COVID-19 and gives a more normalised view of their prospects.

If we are to assume that your standard mature infrastructure company should trade at roughly 15 times net operating cash flow, we can see a few trends starting out. Firstly we can see that both Sydney Airport and APA Group are priced at a discount to a P/E of 15. With Sydney airport having an earnings growth rate of minus 2.72% p.a. and APA Group minus 2.74% p.a. over the next ten years. This either means a better deal for the investors by way of appreciation or the market expects them to trade lower for longer, lose business over time, or drop profits. On the other hand we see Transurban and Spark Infrastructure trading above their respective marks. This implies that, in order to grow earnings and get to a P/E of 15, Transurban needs to maintain earnings growth of 6.25% p.a. over ten years and Spark Infrastructure needs to achieve 10% p.a. over the same period.

As a point of comparison, for NextDC to reach that 15 P/E, their operating cash flows would need to grow at 22.05% p.a. over ten years. We will touch on in a moment why that might be difficult but needless to say, if the first 5 years are at 3.4%, the last 5 years growth needs to be at an eye watering rate of 41.81% p.a. The math looks like this; ($0.118 x 1.034^5) x 1.41806^5 = $0.80 with $12.03 / $0.80 = 15. So amongst its peers of infrastructure companies NextDC appears to have a very high set of future growth rates factored in.

A brief history on predictions

We should start this with a look back to an earlier part of this piece. NextDC shares are priced to have the share price stay stable and the operating cash flows increase by 22.05% p.a. for every year over the next 10 years. If that seems like a heroic effort to you, it probably should. A recent report demonstrated that, even after adjusting for a COVID-19 uptick, data centre spend would reach $6b in Australia by 2025. The interesting part of the article is the assumed growth rate. It comes in at 3.4% year on year. For both growth rates to be true then NextDC needs to not only grow at the industry average rate but also capture a massive amount of additional market share from its rivals. As we discussed earlier, this isn’t a company with small rivals and there are new ones being added all the time.

What would happen if the share price followed the ‘baked in’ earnings growth rate for the next ten years? This part involves some math (lame, I know) and an understanding that the market would only readjust to a price to earnings ratio (P/E) of 15 (assumed industry average) after 2030. Now, with those admittedly wild assumptions in place (perhaps not so wild if you look at the multiples people seem willing to pay for growth) we can look at the share price. Before looking forward it is perhaps intrusive to look back. If you had the foresight to invest in NextDC at the IPO you will very likely be a happy person indeed. Your shares have crept from $1.71 in December of 2010 to $12.03 this month. Now, not for nothing, that’s a 7.05 (12.03 / 1.71 = 7.049269) times growth. A seven bagger (God I hate that phrasing), with a compound annual growth rate of 21.57% ((12.03 / 1.71) ^ (1/10) - 1 = 0.215666).

Well, based on that assumed operating cash flow growth rate of 21.09% ((101.69/15)^(1/10)-1=0.210930), where the current net operating cash flow to share price is 110.06 times, then the future value of NextDC is predicted to be very large indeed. In fact, all other things remaining equal, that would make NextDC the 12th biggest company on the ASX at $40.2b. This would be one step ahead of the $37b money printing machine that is Rio Tinto and you would throw in the value of Lynas Corporation for good measure. This might seem like a strange argument to make but it needs to be looked at within context. Over the last 10 years the share price, as shown earlier, has grown at roughly this pace (21.57%) while spending on plant and equipment (the mainstay of a data centre company) has also accelerated at 24.59% p.a. since 2014.

Without going into an economics/finance 101 lesson, NextDC would also need to capture this market share without dropping their prices to attract new customers because, well you know, money would start to get a little tight. The problem with being in an industry that already has a series of deep pocketed rivals is the only metric they end up competing on is price. They become totally commoditised and start to cannibalise each company within their sector in search of growing market share. This turns out to be great for the customers but not so great for the data centre operators. In the next section we will introduce you to the major in sector competitors but it bears repeating; the growth rates needed to fulfil the NextDC shareholders dreams of riches are, if nothing else, bold.

This serves as the perfect time to speak about the difficulty people have in predicting the future. As Yogi Berra would say “It’s tough to make predictions, especially about the future.” However, we as a society seem to have a pathological need for future predictions, projections, and assumptions in order to execute on almost anything. The end result is a series of well paid professionals, who often call themselves economists, making educated guesses about the future demand of anything and everything. If they get it wrong, they blame the model, or ‘extrinsic factors’, make another guess and continue on their way. Second to weather forecasters (who always get a bad rap) it has to be the only job where you can be wrong most of the time and keep your job for life. For those looking to read further into this slice of modern day life may I suggest you start with Dan Gardner’s book Future Babble.

If you are looking for more recent examples of forecasting done terribly you need look no further than the Reserve Bank of Australia. Now although it is fun to poke fun at the RBA we must do so knowing that the people they employ are, in fact, very very smart. The problem is, as we mentioned earlier, predicting the future is extremely difficult. So every year the RBA would forecast where they saw inflation headed over the next few years. Without fail, year in and year out, the RBA produced a graph that showed the inflation rate heading up and then balancing out nicely. You guessed it, every year the inflation rate undershot the RBA estimates and every year the RBA just had a steeper climb back to their base level. The pattern repeated itself for years without anyone in the RBA thinking that, perhaps, they were starting to look silly. All that is to say that even our country's brightest minds find predicting the future difficult. I can only imagine how much harder it must be for coked out investment bankers who are trying to flog shares to the general public. Oh wait, I can; the bankers just lie. They don’t even care if the company doesn’t hit their growth rates, they only really care about the fees they’ll earn and the end of year bonuses.

One last example of predictions I’d like to finish with was taken from Trevor Sykes fabulous book The Money Miners. When a temporary supply side disruption occurred in the world nickel market, the price of nickel went to the equivalent of £118,000 per ton (today it is $US16k). Poseidon Nickel right at that moment happened to find a nickel deposit near Laverton in WA and their shares took off. This was within a 1969 resources bubble more broadly but Poseidon was, by far, the biggest beneficiary. In September of 1969 their shares were worth $0.80. By February 1970 they hit $280. Then, lo and behold, supply came into the market given the huge nickel price. Also, it turns out that the deposit Poseidon found wasn't as big, high yield or as easy to get out as everyone thought and... they went to receivership in 1974. With the benefit of hindsight it is easy to see but at the time everyone investing in Poseidon on the way up thought they were doing the sensible and profitable thing. Predictions after all are hard.

Does NextDC have the whole market to itself?

In a word; no.

Equinix, a US$65b listed company that last year generated US$5.5b in revenue, is a data centre specialist that is competing head to head with NextDC. Another American challenger is Digital Realty, a US$40b company that took in US$3.2b in revenue last year. Lastly is the $1.13b Macquarie Telecom which took in $266m in revenue last year. These are just the big hitters with many more upstarts hot on their heels. In fact, on the 10th of November another competitor, DC Two, listed on the ASX. To say there is competition in this sector of the market is probably underplaying it.

This ramp up in competitors is exactly what you would expect to see in a red hot market segment with tail winds pushing it along while the jet fuel of ultra-loose monetary policy ensures everyone gets to be part of at least one IPO. What do you think happens next when an industry is on fire like the data centre one is now? Everyone piles in for more, especially when growth is rated at a premium to income in an ultra-low interest rate environment. The data centre land costs start to go up. The components used within the data centres goes up. The staff used to keep the show running see an appreciation in their salaries (NextDC directors, I’m looking at you). I’m fine with all three of those things as a citizen of a capitalist country but it does give me pause. What happens as expenses are going up but revenue is going up faster? If you guessed nothing, you’d be right. There is little incentive for management to crack down on costs when times are really good. That is a task for the rainy day CEO and nobody wants to be that guy or girl.

Correcting the publication bias

In scientific, as well as other journals, research with positive findings is reported on with three times the regularity in comparison to research with the exact same structure, method, limitations, cohort, but the outcomes that are negative or inconclusive. This is called publication bias. It permeates every level of the publication process. Researchers who do the work but don’t get positive results don’t write up the work. Journal editors view all written up reports and include mostly articles with positive findings. Readers spend their precious time reading results with positive outcomes. All of these things come together to see a definite trend towards positive results publication bias.

We, as a society, need to start correcting for this by ensuring that all research, with a positive or negative outcome, is reported on. Now I’m a scientist but I can appreciate that we need more examination of companies than just looking at the assumed future winners. This piece was difficult to write. It isn’t easy putting pen to paper and telling people why they might have invested luckily and not smartly. It isn’t fun telling people they should invest in an index fund instead of a superstar ‘tech’ company. It is however important that someone does it, almost especially after a dramatic run up in the share price.

This is not meant to be a hit piece on NextDC. They appear to be a very well run company with a decent future that involves a lot of organic growth. The real question doesn’t relate to the company as much as it relates to the value of their shares. At the current price of $12.03 you have to assume that a lot of future growth is priced in. In fact, as we saw earlier, this company needs to grow at a pretty decent clip just to tread water with its latest market capitalisation.

The other side of that very dull coin is this; you might be backing the next big thing, but I hope that I've demonstrated that winning might not bring you Amazon levels of wealth. NextDC is, after all, an infrastructure play. Their rate of return will be reliable and they won’t blow up your portfolio, but they won’t send it to the moon either. Eventually the rest of the market will realise the net cash returns are in the ballpark of 4% and will price the business accordingly. If only their CEO was Elon Musk; then the sky would be the limit because, hey, he would tell you so…

Marc Cohodes, a world famous short seller, looks for either frauds, fads, or failures when he looks to short sell companies. I’m not, even for a moment, suggesting that NextDC is in line for a short sale but you have to admit, it does go scary close to being in the ‘fad’ category right now. It is a confluence of factors for NextDC that are largely not of their own making. Tech is big. Just look at the S&P 500 and the S&P 500 minus the tech stocks and you will see how much of the heavy lifting is being done by the tech companies right now. Interest rates are at, or near, rock bottom and central banks around the world are buying up billions (trillions?) of dollars’ worth of bonds in order to push cold hard cash back into people’s hands. Trillions of dollars of investable money is also looking for places to grow right now with the idea of getting a decent income steam almost an afterthought given the low interest rate environment we are in.

I’ve shown you a select few examples viewed with the 6/6 (meters as we are Australian. Embrace the metric system, 20/20 is dated at best) clarity that historical events afford us. It is still too early to tell who will win the data centre wars and I would only caution that you not go all in on one company. Spread yourself out a little bit and lay off some of that unsystematic risk by picking up an index fund. You can know in your bones, just as those before you have, that a new technology is going to take over the world and still back the wrong horse. Or, because you can take the boy out of the country but not the country out of the boy, you can have the right farm but the totally wrong cow. The smart move from here might just be to buy the whole farm.