Month: June 2015

Why I think Andreesen Horowitz is wrong

Interesting Fortune article on a presentation by Mr. Horowitz (a famous VC investor) in which he argues that, despite what to laymen (like me) might appear excessive valuations and activities in the tech space, there is no need to worry about a bubble.

The full presentation can be found here and is very interesting. Although I am low-tech and certainly not as competent on the matter as him, some charts triggered an immediate reaction and I feel the need to comment on some of his slides.

Let’s start with the first one which is supposed to show that (in inflation adjusted terms) we are a far cry from the mania levels of 1999,


Now, I won’t go into the topic of CPI measurement problems, but when I saw this chart I immediately was reminded how, a few years back, gold bulls were arguing that gold, although already in a decade-old bull market back then, was no bubble since it has not taken out its old, inflation adjusted high. (Btw. although I am an adherent of the Austrian school, I have never owned gold as an investment.)

Here is the chart I am talking about,


The fact that the inflation adjusted high was not taken out has not prevented the gold price to fall by almost 40 percent since its 2011 high (the chart would have to be updated). I hasten to add that this has happened despite the fact that the central banks have continued to print money and the deterioration in the geopolitical and economic backdrop all of which are believed to be fundamental to the gold price…

Now, Mr. Andreessen could argue that the fundamentals of gold pricing are debatable (and I would agree), whereas for corporations it is doubtlessly earnings that matter and on this front there is equally no need to worry. He supplies this chart to back up his claim,


No question, forward multiples have already been higher; but nobody is predicting a draw-down of 80% like at the beginning of 2000 anyway – however, trust me, being down 40-50 percent is no fun either.

What we need a sense of perspective: thus, I compare the forward multiple of around 20 with the trailing PE ratio, i.e. the one based on what has been earned in the past 12 months. This allows us to detect potential over optimism in analyst forecasts (not unheard of). Thus, I did a bottom-up calculation (based on individual PE and current market caps) and arrived at a trailing PE of 26 for the index. If you think that (low PE) heavyweights, such as Apple and Microsoft unnecessarily distort the index, you will prefer the average trailing PE of 30. That’s not cheap!

(Note: the Bloomberg trailing PE ratio for the index is way lower at 19.3 as it is based on index weights, which I think is (heavily) misleading here. Horowitz himself seems to use either a market weight or average concept for the forward PE in his presentation which is more meaningful for the question at hand (the index forward PE would be around 16, source: Bloomberg).

With implied earnings growth of between 30-50 percent, the question of who is more bullish, Mr. Horowitz or the average Wall Street analyst, is open for debate. In my experience, this particular forward PE should be taken with a grain of salt.

But there is more, for instance this lovely chart,

Techasapercentage of S&P

Of course, flatness is a matter of definition, but an increase from 15 percent in 2002 to a 20 percent share is not that stable in this context, if you ask me. Further, as every market participant knows, the dotcom mania was different from the current exuberance, in that back then, the froth was concentrated in tech only whereas the rest of the market was actually cheap – a fact which has allowed many a long only hedge fund to sport good risk adjusted returns (and low beta) thereby giving the (wrong) impression of not being overly exposed to the market.

This prompted me to ask the following question: what would the share of the S&P IT index have been in 99/00, if the rest of the market had been valued at the same multiple as nowadays?

Let’s see…

First we need to back out the valuation of the rest of the market at the top, i.e. year-end 1999. My Bloomberg tells me that the trailing PE for the S&P 500 IT sector was 65.43, whereas the S&P 500 (which includes the IT sector, mind you) showed a PE of 29.32. Applying Mr. Horowitz’s share of 30 percent and solving the equation, it turns out that the S&P ex IT was valued at a trailing PE of around 13.8 (!) back then, that’s cheaper than the market valuation at year end 2008, for instance, just sayin’…

Next, I try to determine what would have been the relative weight of the IT index, had the rest of the market been valued at close to current levels, ceteris paribus. I will use a trailing PE of 19 (vs. 13.8 historic) for that matter. A little algebra leads to the interesting result: the IT index’s weight would have been 23.8 percent!


That’s very close to current level of 20 percent and crushes Mr. Horowitz’s thesis (who I think agrees that dotcom was a mania). It further shows how misleading the inflation adjusted chart really is.

At this point I have to add a very good argument made by the author of the article,

(…) While it is true that today’s tech P/E multiples are closer to the early 1990’s than the late 1990’s, a big part of that is because so few of the unicorns have gone public. So is it really an apples to apples comparison? Moreover, while it is true that public equity investors were the main beneficiaries of stock price run-ups for companies like Amazon AMZN2.22% and Microsoft MSFT-0.69%, they also were the primary losers when a wider swath of VC-backed tech companies collapsed in late 2000 and 2001. When trying to understand the current lack of unicorn IPOs, one common explanation is that these companies, while typically older than 1999 IPO issuers, share traits like unprofitability and immature corporate governance. What if VC firms like Andreessen Horowitz are nearing the same type of inflection point that public equities investors hit in late 2000? Different but the same? The flip-side of hoarding value appreciation by keeping companies private is that you’re also increasing risk (…)

Good point!

Anecdotal evidence and personal experience certainly suggest that the bubble in the private markets is even larger than in the public domain…

Then this chart,


Mr. Horowitz tries to calm worries by showing that the so-called “unicorns” (companies with billion Dollar market caps that are expected to be “disruptive” and without earnings) are only a drop in the ocean of tech companies.

Now, while it is true that the “unicorns” pose, ahem, specific challenges to valuation and thus require a large degree of faith, my scepticism is by no means limited to them. For instance, it can be argued that facebook is also kind of loftily valued (trailing PE north of 80 and forward PE of 50+), especially since actual earnings growth has left something to be desired recently.

Or, take amazon (a company which I love as a client): while already mature (by tech standards) and definitely disruptive to all kinds of industries, the proof that it can turn a profit to support its valuation is still open to (heated) debate. While I have no doubts that Jeff Bezos could increase its bottom line by increasing the amount it charges on shipments by, say, a Dollar or two, it is questionable whether orders would not drop precipitously, if he charged more. And since nobody out there can possible know the answer to this question, it can be said that amazon’s valuation also requires “unicorn” like faith, to some extent. It is not for no reason that the adjusted sector weight is close to dotcom levels as demonstrated above.

Further, the valuations of a lot of these are not independent, as there is a feedback loop between, say, facebook’s valuation and the value of every other network/data driven concept. I would like to know from Mr. Horowitz how often he uses comps like market cap/user in VC financing rounds, to show how those revenue free businesses are worth something after all. I bet it’s a lot.

Let me be clear: Feedback loops are normal in valuation, but they are much more pronounced for companies where valuations cannot be supported by earnings and cash flow alone…

No discussion of the tech sector, would be complete without growth opportunities. I found this slide instructive,


First the obvious: I think we all have to agree that the average 1999 dotcom investor  got the growth fundamentals right, as the number people online has increased more than six-fold since the turn of the millennium. Nevertheless, the IT sector index has gone nowhere since then, with heavy volatility in between, proving once more that in the long-run valuation matters.

I note, however, that the bulk of the growth is already behind us, i.e. the effect is already in the (accounting) numbers. For, projected growth of online users is a mere 5 percent annually until 2020 – decent, but I feel the market expects more. Further, it is entirely possible that this growth in users will not translate into commensurate revenue growth (in aggregate) curtesy of falling prices for end products/services.

Mr. Horowitz knows that 5 percent growth is nothing that gets Silicon Valley excited, so he focuses on smartphones where user growth, according to him, can be expected to be significantly larger due to the convergence process between people online and smartphone users. By 2020 everyone who is online will also own a smartphone, he predicts.

As I mentioned in the introduction, I know nothing about the tech industry; I do not know how reliable these statistics are or how the variables are defined and measured. However, I do know – and I am certain this is true for most of my readers – that everybody and his dog around me already has got a smartphone suggesting little room for convergence. The same is true for most emerging markets I have visited. See for example what the situation in China looks like,


There are now already 550 million (!!!) Chinese who access their internet through their mobile devices (I assume this is mostly smartphones). That’s almost twice the entire US population (light blue line). How many more smartphones does China need in Mr. Horowitz’s opinion?

As I already suspected: a lot of the beneficial effect of internet/smartphone penetration are already in “the numbers”. It is very likely that the future will be dominated by the cannibalizing fight for market share.

No discussion of a distorted market would be complete without discussion about the “suckers”, i.e. the dumb money that can be persuaded into buying at what insiders know/feel is ludicrous pricing. Indeed, many observers of public markets (stocks, bonds) do not think we are in a bubble yet, precisely because they cannot clearly identify the naïve market participant (hint: he/she has a PhD and despite his unlimited buying power, no experience whatsoever in practical decision making).

Because in the last tech boom retail baby boomers were the marginal buyer, many are waiting for the retail investor to come back for technical confirmation. Since history does not repeat, but merely rhymes I think they are waiting in vain: more than a decade of below-average interest rates have squeezed the middle class who is strangled by record high taxes and private debt, with no money left for “investment”.

So much for the public markets.

What about the private (mostly tech!) markets where Dan Primack, the author of the article, detects even larger signs of froth?

From the presentation,


When I read “non-traditional investors” I felt relieved and astonished that the sucker theory is so robust – Q.E.D.


Andreesen Horowitz presentation is great becaue it gives us an excellent introductory overview on the tech industry. Going through the slides (there are many interesting slides I haven’t discussed) you get a sense of what moves and motivates the key players in the field.

However, my reading of the slides confirms my view that valuations in the sector are lofty and expected growth likely to be way overstated. I think there IS a bubble in tech (as well).

 (Disclosure: long MSFT and CSCO)


Nothing will happen with Greece this weekend

As you are aware, the situation in Greece is not getting any better. This chart shows the evolution of Greek bank deposits over time,


This is ugly: deposits have fallen substantially since November when I had my discussion (here) with David Einhorn about Piraeus bank. This should add substantial funding pressure for the banks. Obviously, it is out of question to “grow your loan book” in such an environment.

One of the arguments we had been about the cost of funds, here is what David had to say,

(…) The 2.04% is the cost of new time deposits in August 2014. This is not the average cost of deposits in the Q2 2014 financial reports, which was 2.62% according to the same presentation (and 2.71% in Greece time deposits). This cost will reduce down to the new cost of deposits over the 1-2 quarter duration of the time deposits. So even without any additional lowering of deposit pricing you would still expect to see a near 60bps improvement. This is obviously very material given there are €29.6bn of time deposits (the rates are also coming down on savings and sight deposits but this is much less relevant). The 60bps is worth close to €200m of additional annualised PPI, compared with the €1.2bn annualised reported PPI in Q2 2014. How low can deposit rates go? Time will tell of course but rates in the other periphery are around 1% and the management teams of the Greek banks are clearly trying to reduce them towards this level. I would suspect another 50bps (…)

So the average cost of new deposits was 2.04 percent in August 2014. Out of curiosity I looked up deposit rates Piraeus bank currently offers clients who want to open an account (here).

And what do I see?

Rates range from 0-1.4 percent – lower than in August 2014!! (I note that they offer a multitude of complex products, presumably to con their clients akin to mobile companies).

To sum up: a substantial drop in supply (of deposits) has not caused any increase in the price of funding. Given that the balance sheet of the Bank is more or less unchanged, we can rule out a commensurate drop in demand (for funding).

Do we need to rewrite economic theory? Is this some puzzle, something which has an unorthodox answer? Maybe rates stay low because there is a lack of profitable investments in the Greek economy? Maybe it is inflation expectations?


Rates are low because Piraeus bank does not have to bother about funding. There is a (very) large market player who does not make economic decisions and happily supplies any amount of funds needed. Market participants have an expression for that: dumb money. You know who I am talking about,


Deposit rates in Greece are low because the ECB (technically the Greek central bank) provides any amount of funds needed. There is no need for Greek banks to attract funds or keep costumers.

Now, the interesting thing is that any day that passes with deposit outflows (and higher ECB credit lines) the negotiating position of Varoufakis and Tsipras (who have made ludicrous promises to the electorate) improves. Why, that’s what I wrote in another post in February (here),

(…) Or take the ECB. A write-off would finally make it clear that it has engaged in government financing, although explicitly forbidden by EU contracts. This certainly would mean the end of the fiscal union debate. As someone who is tired to see government economists and “quality” newspapers argue for that nonsense, I would highly welcome such a development – even if I am aware that it would lead to a big economic and political crisis. (I confess that am a fan of the German proverb: “Besser ein Ende mit Schrecken, als Schrecken ohne Ende!”)

And this is exactly the reason why a write-off will never happen, as long as extend and pretend works. The accounting of government is opaque for a purpose. It has never been otherwise. If you do not believe me, read this book, it will cure you of any illusions you might have (…)

That’s right, even though the Eurocrats certainly dislike the two Marxist lunatics who make fools out of them, and even though the Greek economy (or the fate of ordinary Greeks) do not matter at all to them, they try to avoid a (formal) default at all costs, otherwise there is a rude awakening in the electorate in the core countries. And the costs are increasing by the day…


Although many predict this weekend could bring capital controls and the like in Greece, I do not think anything will happen. Instead, we will see further extend and pretend. Do not forget that a default would come at a very inopportune time now that the (of course independent) EU top court has decided that the purchase of government bonds is not government financing (here). This is probably a bold statement given current headlines, but I will stick to it.

Let me be clear: I am NOT predicting this can or will go on forever, or that the parties involved can “solve” the Greek problem. Surely, at some future point in time some kind of shock will occur, I just do not feel it is time yet. And aren’t market crashes more likely to happen in autumn anyway?

Excellent Paper on Chinese Capital Flight

Reader Nick responded to one of my last posts and recommended an interesting paper: „High Wealth Concentration, Porous Exchange Control, and Shocks to Relative Return: the Fragile State of China’s Foreign Exchange Reserve“ by Victor Shih from Northwestern University.

The author argues that China, despite high FX reserves, is vulnerable to capital flight due to the fact that a large share of its wealth is concentrated among the top 1 percent. A shift in investment preferences among this group has the potential to dangerously deplete this stockpile despite its huge size. Shih thinks that investment considerations of the Chinese elite are mainly driven by the real interest rate in CNY and at the time of writing (2011) the favourite conduit to escape an erosion in purchasing power was Chinese real estate (this certainly has changed with the housing slowdown since then). Confirming my analysis from a few months back (see here) capital controls are a non-issue for the well-connected and numerous ways exist to circumvent them, most prominently fake trade invoicing.

Interestingly, Shih has an inconvenient message for those who think that the potential for financial repression are unlimited for China’s government,

(…) In a repressed financial system, ordinary households must suffer in silence while inflation erodes both their income and savings. However, the wealthiest households in China, which command a sizeable share of China’s household deposits, will look for ways to beat negative real returns in the banking system (…)

Not so easy to repress the well-connected…

But it is Shih’s account on how the Chinese authorities might respond to capital flight that’s interesting. So interesting, in fact, that I have decided to quote in full with comments in between,

(…) If the FX reserve declines by 50-100 billion a month for a couple of months, the Chinese government will declare victory in their rebalancing efforts and signal the end for yuan revaluation. This moderate amount of net outflows would be an unnoticeable blip in China’s enormous FX reserve. There is some chance that smart money that smart money may see this as a sign of coming trouble and may accelerate their reallocation overseas. If the FX reserve declines by over 100 billion USD a month or if it declines two months in a row, the Chinese government will try to hide it by having China Investment Corporation (CIC) repay some bonds to the PBOC before maturity with US dollars (…)

Currently outflows are coming in at about USD 30 billion/month (see here), we are still away from real worries for the Chinese government. However, I will look out for headlines involving the CIC,

(…)This would force the CIC, which is separate from the State Administration of Foreign Exchange (SAFE), to first sell its equity or bond holding to raise money. Because over half of CIC’s 330 billion dollar holding is tied up with Chinese bank shares, the CIC cannot quickly sell a large portion of these holdings without causing a crash in the shares of Chinese banks, which would intensify a financial panic (China Investment Corporation 2010)(…)

The paper was written in 2011 so AUM are probably much higher now. The market would try to front run CIC which makes bank shares a potential leading indicator of CNY devaluation,

(…)Even this amount of FX depletion would be hardly noticeable and is well within SAFE’s ability to control. If there is net outflows exceeding 250 billion dollars within half a year, SAFE will try to stem the flows. Instead of carrying out inspections on export invoicing, SAFE will carry out inspections on import invoicing (…)

In other words, watch out for headlines about a crackdown on import/export invoicing, it may indicate increasing capital outflows and give you an edge before the official numbers are out. The following is also interesting,

(…)There will be a crackdown on underground money changers also. Indeed, when faced with an unusually high amount of outflows in the second half of 2008, SAFE began to carry out inspections against outflows (Wen and Huo 2009) (…)

Put differently: there is a precedent for what will happen once USD liquidity starts tightening globally. It also indicates that the Chinese malinvestment boom would have deflated much earlier had the Fed not engaged in large-scale QE…

(…) However, just as the prevention of hot money inflows is ineffective today, the attempt to stop capital flight will be challenging. China cannot stop importing large quantities of commodities and food from developing countries, especially if there is inflationary pressure. Chinese import of foreign goods totaled over 1.2 trillion USD in 2010, a 40% increase from 2009 (National Development and Reform Commission 2010). As such, the Chinese government cannot completely stamp out import invoicing as a channel of capital flight (…)

The power of the Chinese authorities to “manage” the economy is way overstated. Once the tide turns, they have only limited options.


It is likely that I have been too early with my CNY short. The political elite who controls most of the wealth in China is likely to get to know first about an impending devaluation and would try to move its assets out of the country as fast as possible. The Chinese authorities know this and hence are probably reluctant to even discuss this as a possible solution. The CNY will be last shoe to drop.

Chinese bank shares are probably the better way to play this (apart from the AUD), as CIC reserves consist to a large part of bank shares. In case of continued outflows there would be selling pressure similar to what occurs when mutual fund holders redeem their shares. Moreover, with the banks you have another lever, since they must have considerable credit risk given their non-existent lending policies (how else, for example, do you explain this). Shorting a basket of Chinese bank shares with long-term put options could be an interesting play.

Disclosure: short CNY and short AUD