Why Deep Work is essential for keeping up with robotics (career and investment opportunities)

Executive summary: AI is coming. You’d better think about your career and investment opportunities in AI and robotics. And the importance of Deep Work to keep up.

Hint: Google, FB, Amazon

Length: 2128 words


The robot overlords are drawing closer

Finally a machine beat a decent player at the ancient eastern game of Go. With decent I don’t mean an Asian grand master, but at least a three time European champion.

google deepmind

Artificial Intelligence keeps progressing, no matter whether you know about (or like it) or not.

First an AI application is typically seen as a curiosity. Then it becomes a tool you need to learn how to use. And eventually it will develop to the point where it could take your job.

IBM’s Watson easily beat the world’s best Jeopardy masters several years ago. Since then it has become the world’s foremost oncology expert.

Currently Watson is on its way to start replacing swathes of paralegals at law firms as well as finding new oil reserves. A few years down the road, anyone with a cellphone (or AugReal contact lens) will be able to tap into Watson-like powers for any kind of search or research.

IBM Watson onc 2

Over the coming 20 years, most jobs will be affected by the progress in robotics and AI. Hence, even if you are aiming for future-proof industries (The 5 Singularity Enablers or The Big 5 human issues, or, most likely, a combination*), you’ll nevertheless need to learn to work with robots and artificial agents, or risk replacement.

*I have mentioned the 5+5 in earlier posts on job security nr 1 and nr 2 and my post about programming: nanotech, biotech, AI, robotics, additive manufacturing and energy, water, pollution, food, longevity/health

 

Professions at risk

If you’re flipping burgers, sewing garments, assembling consumer electronics, building cars or houses, reading court cases, writing (sports) news, trading stocks or driving any type of vehicles for a living, you’ll soon be out of work (except if you can leverage the new technology in some creative way).

 

Deep Work makes you change-resistant

To stay one step ahead of the AIs; to be a fast learner as well as able to tap into your most creative powers, one indispensable skill for the future will be the ability to perform deep, focused thinking and problem solving, i.e., Deep Work, in the words of Cal Newport.

Rather than allowing various notifications from e-mail, Twitter, Facebook. Instagram etc. to force your days into shallow, responsive, always on-line type of activities, you should practice going off-line and “deep” for longer stretches of time (30-90 minutes) as often as you can.

deep work cal newport

Buy it, it may be your best investment ever

Deep work restructures your brain, making it easier going forward to enter a state of flow and focus, and thus becoming more and more effective, and increasingly able to perform above AI-level, not to mention quickly learn new skills, including how to use new (AI) tools.

Meditation helps too, but that’s just too weird and Eastern for most – at least if we’re talking about 30-minute long sessions or more. I myself had much rather turn everything off for 90 minutes and solve an intellectual problem at the top of my ability (writing a blog post, a new book, or working if I had a job).

 

If you can’t beat the robots, join them

This Tuesday I (and my dog Ronja) talked in front of 400+ engineering students at Sweden’s top technological university, Chalmers. I specifically remember three interesting questions.

  1. What made your fund so successful?
  2. What advice would you have given yourself as a student today?
  3. What would you have studied today?

ronja chalmers

Answers to important questions are superpositions of the entire spectrum of answers

What’s particularly intriguing about the first question is that my number one success factor was also the number one negative at the same time:

Being unbiased and fundamental

It worked extremely well over the several bull and bear waves during my career (1994-2015). On the other hand, being unbiased and fundamental also made it all but impossible to ride the bull waves (“bubbles”) long enough – in particular the last one (the quantitative easing bubble that finally seems to be bursting).

I often find that the most important questions are answered in the same “how long is a piece of string” fashion. Learn to recognize and acknowledge those situations, rather than dislike them. They provide you with a much wider range of choices than clear-cut 1 or 0 situations.

 

Have fun no matter what you do (Sinéad O’Connor)

Nr 2 is also easy: Direct your studies and career toward something that interests you, that you like, that gives you the “unfair advantage” of having your favorite hobby as your work.

No matter how successful (or not) you get, at least you had fun on the way. And, given the unfair advantage, you’ll probably change the world in some positive way and become rich/famous/of stature, no matter if you want to or not.

Build authority through usefulness (for you and others), not (empty and meaningsless) celebrity or wealth

Sprezza and Ronja at Chalmers

I however began by saying that I myself would never have listened to such advice, and that I didn’t expect anybody to do it today either. At 22 I was too focused on making money, or at least on getting a job, any job. In my eagerness to become independent I as quickly as possible put myself into wage slavery and the consumption rat race.

In my defense, I hardly understood the concept of starting a business. I had a very static view of the world and kind of thought all companies already existed.

Hmmm, that didn’t make me look any better did it?

 

Resistance is futile

The third question could have been answered in a myriad more or less complicated ways. To keep it simple and clear, I boiled it down to one single word: Robotics.

robot sprezzaturian

Going east, to Japan, China or South Korea, is preferable for anybody going into robotics. However, it probably isn’t necessary – and France, Germany, The U.S. and Sweden also hold their own within robotics. It’s even possible Google and other U.S. companies are on their way to overtaking the Asians.

Except for North Korea of course. Their mighty leader has already built a super strong general AI that will rain fire over the western subhuman devils. Just recently, e.g., Kim Jong Un’s AI crafted the most efficient and gloriously superior hydrogen fusion bomb that history will ever see.

KJU

I wonder if he has any idea how funny he is

 

Why robotics?

Because everything comes together there. And it’s the most future-proof industry there is.

If you want to get really dystopian, in the future the only humans left are the ones tending to the robots. In a parallel universe, robot owners and robotics stock owners are the ones holding the upper hand.

 

The geeks shall inherit the earth

 

Robotics is industry’s equivalent of Deep Work. Every single part of a robot is developed at the leading edge, at the top of everyone’s ability; technology, biology, biotechnology, neurology, philosophy, psychology, programming, materials, motors, artificial intelligence. The combination of these into useful and robust machines demands even more of the creators (i.e., you).

If you are wary of being made redundant by automation (and you should), the obvious solution is to be the one controlling the automation (and inheriting the earth).

 

It’s not “just robots”

There is so much to do in robotics: vision, balance, appearance, movement, safety, reasoning, emotion, interface, power source, touch, control. And each of those need to be craftily integrated with the others, e.g. vision, prediction and balance.

robot worm sprezza

Take just vision as one example. Ideally you’d want to combine radar, ultrasound, stereo vision, texture analysis algorithms, laser, object data bases, blueprints, recent memory, inference algorithms etc. in one single system, in order to rapidly and reliably map the environment in 3D, as well as make forecasts for the coming milliseconds, seconds and possibly minutes to prepare probable movements.

Similar issues are facing research teams within, e.g., balance/movement (motors, artificial muscles, scenario simulation, limb synchronization. machine learning) as well as other important sub-segments of robotics.

At the user end, there really are no limits to where robots and AIs might go: industry, care, household, status, services, shopping, news, education, search & rescue, surgery, research & exploration, sex/porn and on and on. Every single area of life and business will be affected in the coming two (or maybe three) decades. Read more about the steps to AI here, or read, e.g., Kurzweil’s book “How To Create A Mind”.

Simply put, robots is where all technologies converge. It’s a place to perform deep, accelerated, learning and highly value added work. It’s a place where cross-discipline knowledge and deep, associative, and lateral thinking will come at a premium.

My advice is to take your favorite area (hobby), and combine it with some supportive technology and apply those in robotics. I can’t see how you could lose.

If you are an engineer, or psychologist, or designer; if you go into robotics and AI, if you take the “unfair” angle of doing it from the standpoint of fascination, you just can’t lose.

I may have told these guys at Chalmers that aiming for an employment at a big engineering firm, rather than starting your own business, would be just sad…

chalmers

On the other hand, I also told them they had no business worrying about the coming recession or not earning their livelihood. Very few, if any, of them will even notice the coming recession – except for from tabloid headlines.

 

Focus, go deep, and go robot

Don’t forget my message to prepare for a fast-changing world by practicing true, off-line, no notifications, focus, as well as perhaps playing Go and meditating, lest constant e-mailing every day will erode your protective myelin layers in the brain completely, making future deep work practically impossible.

Without the ability to perform deep work, even going into robotics won’t save you in the long run. So, leave the 99% behind and commit to stretching that brain of yours regularly.

But now on to something completely different, investing in robotics and AI companies

 

Investment opportunities

Sorry, there are no free lunches.

Well you didn’t think I would hand out free buy recommendations, did you?

handing out free stock recommendations

Anyway, here are some AI (and robotics) related companies to think about.

I’m not saying you should buy them (now) or sell for that matter, but they are definitely worth considering at the right valuations.

Since I’m sure you can come up with more companies, and more pure plays at that, I hope you’ll put those in a comment for all to share.

Top 5 most obvious AI companies

  • Alphabet (Google)
  • Facebook (M, Deep Learning)
  • IBM (Watson, neuromorphic chips)
  • Apple (Siri)
  • MSFT (skype RT lang, emo)
  • Amazon (customer prediction; link to old article)

Yes, I’m US centric. So sue me :)

 

Other

  • SAP (BI)
  • Oracle (BI)
  • Sony
  • Samsung
  • Twitter
  • Baidu
  • Alibaba
  • NEC
  • Nidec
  • Nuance (HHMM, speech)
  • Marketo
  • Opower
  • Nippon Ceramic
  • Pacific Industrial

Private companies (*I think):

  • *Mobvoi
  • *Scaled Inference
  • *Kensho
  • *Expect Labs
  • *Vicarious
  • *Nara Logics
  • *Context Relevant
  • *MetaMind
  • *Rethink Robotics
  • *Sentient Technologies
  • *MobileEye

General AI areas to consider when searching for AI companies

  • Self-driving cars
  • Language processing
  • Search agents
  • Image processing
  • Robotics
  • Machine learning
  • Experts
  • Oil and mineral exploration
  • Pharmaceutical research
  • Materials research
  • Computer chips (neuromorphic, memristors)
  • Energy, power utilities

No, I won’t help you with valuations. However, I just might tell you in my newsletter when I start buying.

 

Summary – deep work and robots

Rapid changes require fast learning of difficult material

Fast and solid learning requires deep concentration

Deep concentration demands time off from e-mail and social media

To prepare for the future, practice focus, use your time in deep focus to advance your skills and understanding of how you can contribute to the field of robotics.

And Go East (Japan, China, South Korea)

Or possibly west.

Or stay where you are.

And think about becoming an owner of AI and robotics companies while there is still time. I plan to buy some of the most obvious ones (including Google) in the ongoing market downturn (2016-2017).

 

Please help spreading this important post

I think this is the most important article I have ever written, and I would like as many students and other interested people as possible to get the chance to read it.

So, please help spreading it as widely as possible through your social networks: Twitter, Facebook, Google+ etc., and tell your friends about it IRL too.

P.S. “Subscribe” If you are new here, don’t forget to subscribe to my newsletter (weekly updates and retarded recounts) and get my free eBook about my 15-year struggle at the European Hedge Fund Of The Decade (“The Retarded Hedge Fund Manager”).

P.P.S. “Glorious future” No, I’m not dystopian. I actually think we’re headed for a glorious future. After a couple of years of economic and financial rout that is.

Amazon redux – this time twice as expensive

Warning, this is a long hard-core finance article. It’s mostly about the valuation of Amazon. The stock is so expensive, I can’t even… My blue sky scenario is -90%.


 

March 2000 was a good year for drunks…

amazon meme

I’ll show with a few simple assumptions and calculations why I think Amazon’s shares are 10-20 times as expensive as I would find attractive.

I’m not advocating going short the stock though. You have to realize when you don’t “get it”. At this point I’m not quite sure what would change my mind about this. It’s just so far off from what I consider reasonable that the situation simply is out of control. Hence, better not risk any money.

All I really do in this article is extrapolate growth and margin trends, and then state what kind of return I demand from an investment to consider it worthwhile.

You are free to have different preferences or assume others (i.e. the market) have.

Read on if you are interested in a step by step pedagogical approach to back of the envelope investing

Beer or IT stocks?

beer darth vader

At the height of the stock market bubble in March 2000, you would have been as well off financially if you bought beer and recycled the cans (in certain states), than if you “invested” the money in many IT shares.

As it were, the entire Nasdaq index (not only tech stocks) declined by 80 percent, and many stocks declined by much more. Just cash would of course have been so much better than beer, not to mention bonds or gold.

The difference between an 80% drop and a a 90% drop is an additional fall by 50% in value.

And yet another 50% loss takes the total loss to 95%. A third sequential 50% loss, after the initial hypothetical 80% plunge, renders a total loss of 97.5%, which aptly describes the trajectory of many IT and new media stocks during the years 2000-2003.

 

Stars do fall; Nasdaq composite index 2002-2015

nasdaq 25 sept 2015

I should know, since I was chiefly (ir)responsible for buying shares in an internet consultancy called March 1st (named after its starting date March 1, 2000, at the very peak of the bubble). Luckily, we made back the initial losses on that position (and sold it before the bankruptcy in early 2001). We then went on to earn so much more shorting similar stocks in Europe, using the US stock market as a template.

Make informed investments, with whatever you have left after Mr Market taught you a lesson – always be investing

The IT crash was no surprise

To many thoughtful and experienced investors the crash of 2000-2003 didn’t come as a surprise. I have a list of 50 prominent investors and other pundits, that intelligently and timely forecast an epic bursting of the insane IT bubble. Many of those went on to correctly predict a housing crash and a financial crisis after 2006.

Neither of the last two crashes came as a surprise to anybody with the slightest knowledge of the art of valuation and the history of markets and bubbles. The timing, however, was often off by a year or two – or more. 

Crash number 3 is in the cards

I think it’s about time again for a correction of frothy share prices. And by correction I mean on the order of -50%.

There are dozens, if not hundreds, of fingers pointing to that conclusion, but this is one of my favorites:

Dr John P Hussman’s chart over “Price/GDP” vs. stock market returns

Observe, e.g, where the blue line bottomed in 2000, correctly forecasting ten years of negative annual returns 2000-2010. Or peaked in 2009, forecasting 12% annual returns 2009-2019 (with the market so far well on its way to fulfill that prophecy).

hussman 14 sept 2015

Chart from here at hussmanfunds.com

  1. Notice the tight correlation
  2. Notice that when correlation falters for a while, it comes back with a vengeance. Actually this “error” in itself makes a better job predicting market returns than the revered “Fed model”
  3. Right now, market valuation points to negative annual total returns, including dividends, over the coming 10 years
  4. Usually the market corrects to levels where you can look forward to 10-20% annual returns over the coming 10 years. To get to that point the market needs to fall by 50-75% or stay level for 10-15 years. Which do you think is more likely?
  5. A vertical move upward by the blue line is tantamount to a stock market crash, thus changing the valuation from promising, e.g., 0% return to promising 10% annual returns the following decade.   

 

Amazon – buy the merchandise not the shares

I  have bought 2 Kindles from Amazon and hundreds of books. It’s a great company – for its clients.

However, let’s take a closer look at Amazon’s shares as opposed to their services. A quick visit to Market Watch tells you the stock price is 528 USD/share, with a total market cap of 251 bn USD.

There is no useful information on valuation, though, (since Amazon operates at a loss and doesn’t distribute dividends).

So, what do you do? Ask your friendly neighborhood hedgehog?

-No, you do the math yourself. If you consider yourself a real investor, that is. But first, let’s find out if you’re cut out for this. What kind of money person are you?

 

Amazon’s share price 2002-2015

Amazon 2002-2015

N.B. Amazon’s stock price has already fallen by >50% twice, and by >30% several times

 

Astute Investor or Ignorant Speculator

  1. Active or Passive. A passive investor buys and holds single stocks or index instruments -and never sells. An active investor buys and sells at various intervals, depending on a multitude of factors (we’ll get to that later)
  2. Single stocks or index investing. If you chose index-only investing, forget about Amazon (duh!)
  3. Fundamental or Technical. Are you willing to make an effort to understand what you invest in, do the math so to speak, or are you more inclined to speculate on tips, momentum, stock charts etc?
  4. Hope or Safety. Will you trade on hype and hope, on blue sky scenarios, or do you have the discipline and patience to wait for opportunities that offer a margin of safety

If you are an active investor and single stocks focused you can proceed. If you are passive, just buy the stock or an index already and go back to sleep.

 

See. Pretty picture

If you just can’t be bothered with math, spread sheets, forecasts and economics, do what so many others do: draw pretty pictures in stock charts and invest according to the patterns you think you can discern.

Or just follow whatever momentum or trend you want to see or you are told (tipped) about by a “friend” or broker. Just don’t blame me when you go broke.

technical

lines in a chart are not real

 

Your broker is your adversary

Stock brokers don’t want you to make good investments. They don’t want to be your friend.

They want you to pay a commission.

When you buy, somebody else sells. One of you is often making a mistake. And the broker couldn’t care less.

Charlatans (stock brokers and company representatives) use extremely simple and misleading measures of valuation such as some future year’s Price/Earnings ratio (given made-up profit margins), perhaps complementing it with an arbitrary, historical growth number.

“Amazon has increased its number of employees by 38 per cent in the last 4 quarters and at ‘normalized’ operating margins of 10% it’s trading at an EV/EBIT ratio of just 21.5 x (12 months forward op. earnings), i.e., only half its (employee) Growth rate! PEG=0.57 which is below 1, hence BUY!”

Assumptions: TTM sales of 95.8bn in June 2015 will grow at the same 17% rate the coming four quarters to 112.1bn. The operating profit at a 10% margin will be 11.2bn. The current EV is 251bn-10bn cash pile=241bn. The EV/EBIT thus comes to 241/11.2=21.5. The growth rate used in the PEG calculation is the last known employee growth rate.

 

Stock pitch break down

The above may sound convincing. But if you look more closely you’ll see the cracks in the facade:

  • Okay, so the number of employees grew by 38%. Even if it could be representative of future growth in some cases, it’s also a cost now. Sales, however, only grew by 17% the last four quarters (despite a 36% growth inemployees in the preceding period), and slowing. Thus, a better forecast of sales growth going forward would be 15%, given current trends.
  • In addition, the relevant growth rate, for applying a valuation multiple, is the growth rate going forward after the year of the multiple (which probably is even less, given the deceleration the last few years). Actually, unless profitability accelerates soon, Amazon needs to taper its hiring pace, which should hurt sales growth even more. Nevertheless, let’s give them the benefit of the doubt for now and assume 15% sales growth. We can refine the model later.
  • “Normalized” profit margins at 10%? Whats “normal” with 10%, when actual margins have been between 0-1%? Are there any plausible reasons to forecast higher margins than what Amazon has produced the last 5 years?
  • Actually, there might be. If Amazon reaches ‘scale’, when growth diminishes, margins could increase. This is the trickiest forecast of them all, since Amazon hasn’t reached “steady state” with sustainable growth and margins, so what to do?
  • Look to similar companies for guidance on potential margins. Also make a sanity check on what ROE those margins imply for Amazon (Return On Equity). Few companies stray outside 5-20% ROE for long. A higher ROE draws in competitors, and a lower ROE is a waste of capital. Assuming Amazon can increase its net margin to 1.5% (operating mgn approx 2%), Amazon will make a decent 14% ROE (1.5%*112bn/11.768=14.3%). Hoping for more is just that; hoping and believing the hype.
  • EV/EBIT-ratio of 21.5. Why is that supposed to be cheap in an absolute way? How can you tell? To start with it should be compared to profit or cash flow growth, which (at steady margins) can be approximated by sales growth – in our example at most 15% (probably less after 2016). Suddenly the PEG is 21.5/15=>1.43 instead of 0.56. That still doesn’t say anything about whether it’s a good buy or not, even if it can be compared to other listed companies’ PEG ratios. Anyhow, it’s apparently a lot more expensive than what the broker first said.
  • And, I almost forgot, that was based on a 10% op. profit margin. Assuming 1.5% instead, renders an EV/EBIT of 143 and a PEG of 10. That is 19x expensive-er than the broker was saying.
    • In addition, whatever happens in 2016 is less worth than if it were today, so profits should be discounted by your required rate of return. An EV/EBIT of 143 a year from now is some 5-10% more expensive than the same multiple today, or >150. That however still doesn’t answer the question whether Amazon’s stock is expensive. Do, however, compare the 150x on trailing EBIT to the more typical 10-20x multiple seen in the S&P 500 index. Enticed?
  • Continuing to use the extreme simplifications above, you can calculate an earnings yield to see what your actual return would be in a single year. If you dare extrapolate that year indefinitely, you can even rely on it as the stock yield. In Amazon’s case, given 1.5% operating profit margin and a 30% tax rate the net profit is 112*0.015*0.7=1.18 bn and its earnings yield is (profit/market cap) 1.18/251 = 0.47%.
  • If you are lucky all this profit generation is converted to cash and distributed to you in some way. Also, if you are lucky, the coming four quarters are representative of the coming eternity for Amazon’s business (in practice only about 50 years count).

There you have it. Given a set of simplifications, you can expect around 0.5% yield on an investment in Amazon. It doesn’t really say anything about what price you’ll be able to sell the stock for in the future, but if you just hold on to it forever, you’ll get an 0.5% annual return on your investment, unless Amazon’s business changes radically.

 

The quick and the dirty

Happy? I’m not. And that’s not because of the low return, or that the stock might (should) fall dramatically in the meantime (to a price where you could expect at least a 7% annual return). It’s because of all the simplifications.

Just because you put a lot of numbers down on paper doesn’t mean they are correct or you are any wiser

 

Also note that, if in the future, when you want or need to sell the stock to get your capital back, another valuation paradigm dominates, perhaps demanding the historical 6% or 10%, 15% or even 20% return on equity investments, the Amazon stock price will be proportionally lower.

At a 9% required return, Amazon’s stock price would be just a twentieth (27.50 USD/share) of what it is at 0.45% required return.

That was the quick and dirty way of analyzing and valuing single stocks. You should at least do as much for any stock you contemplate buying.

  1. Estimate (extrapolate) sales growth and margin trajectory in an informed manner
  2. Calculate future profits and cash flow that take into account total available market, innovation and competition
  3. What yield in relation to the company’s market cap does that cash flow produce?
    1. If you’re happy with that yield and don’t care if the stock falls in price, you’re good to go
    2. If the yield is below your required return (9% for example) or the historical market return (7%) and you do care about price fluctuations, tread carefully. 
  4. Is that enough? Are there better alternatives? Is it enough for others (i.e., the market)?

 

Get out of here!

If you felt that was already too much, you are not an active or fundamental investor. A sell side analyst does a lot more. It’s not necessarily better or more effective but it’s impressive work.

 

The long winding road of doing sell side level research and stock analysis

Make company forecasts like a ‘pro’

  • Make reasonably detailed Profit & Loss, Balance Sheet and Cash Flow forecasts. These are hundreds or thousands of Excel rows long, several hundred columns wide and dozens of sheets thick 
  • To do that you need to forecast sales, costs, wages, investments, loans, interests, taxes, dividends, buybacks etc
    • Any one of those components can be forecast using a dozen other inputs
  • To do that you need a view on demand, competition, currency (FX) trends/fluctuations, technological development etc.
  • You also need the company’s history to have an inkling of how its products have been received by the market and how management has been able to juggle all variables historically.
 
In addition to this fundamental single stock valuation, you should add a handful of supportive layers:
 
  • Total market valuation (if the stock market in general is “cheap” you stand a better chance the entire market will rise and help your shares appreciate. The latter, however, only really matters to a hedge fund and not for a long only index fund)
  • Momentum in the stock or the market (positive market trends can lift all boats)
  • Risk tolerance, i.e., your willingness to gamble (and willingness to lose)

Stock prices move in cycles. Sometimes a certain industry or company is in vogue and sometimes it’s not. You can try to speculate blindly with this momentum or you can choose to invest whenever the price (valuation) is below a certain maximum threshold and sell when it is too high for a comfortable plausible return. Dare to make your own judgement of value and buy when others don’t.

It’s better to look the fool before, than after an event

 

Analysts make meteorologists look good

There is no secret to forecasting. No one can do it anyway and stock market pundits and economist are among the worst forecasters of all.

I think my complete immersion in finance for a quarter century is why I have always marveled at the accuracy of weather forecasting (no irony, I am amazed of what they can do).

The following can serve as a rough guide to informed forecasting of financial data. But remember that your forecasts don’t become the truth just because you put them down on paper.

Collecting historical data

  • Get at least five to ten years of history for a simple set of parameters for the P&L: sales, cost of goods, cost of personnel, cost of loans, taxes
  • Make a rudimentary balance sheet: real Long-Term assets, goodwill and intangibles, Short-Term assets, cash,  net working capital, equity
  • Cash flows: operating profit after tax, investments, changes in Net Working Capital
  • Double check for funnies: too low wages vs history or vs. competitors, high personnel option compensation, share count development; profit, cash flow and equity not adding up over the years etc. Make charts of every series and subseries you can think of an look for kinks, trends, holes, bumps and so on. Then investigate those by reading old Q-reports, ex ante guidance commentary and ex post explanations.
  • Check the latest quarterly trends more carefully: DSOs, e.g., (Days Sales Outstanding=accounts receivable/sales) = how easy is it to get paid, is the company overselling or even sending unsolicited invoices?
 
Making forecasts like a baws
 
  • Find leading data series in daily, weekly, monthly or quarterly data: from competitors’ sales, costs etc, weather data, monthly national data, PMI surveys, national industrial output statistics, retail sales, FX, etc
  • Start forecasting: Basically extrapolate trends in historical company numbers and leading data series. Allow for cyclicalities and reversions to the mean. E.g., profit margins and Return On Capital Employed numbers tend to gravitate to a mean for the industry or the economy. Do not expect super profitability or super growth for eternity. Check how long the very best companies of all time sustained their growth, margins or returns. Check what the average company did, including those that failed and are no longer here. Easiest is to check nation wide numbers for margin and growth trends.
  • Analyze really long cycles (over several decades or longer) and trends to make sure you don’t get myopic, narrow sighted, seeing only one half of a cycle and thinking that is a long term trend to be extrapolated without risk
 

Knowing what you don’t know

All of this is particularly difficult for new companies and even more difficult if they operate in new industries. There is no history, no precedent. That doesn’t mean you get off more easily. Quite the opposite. The risk is higher and your work is harder.

Sell side analysts take advantage of the lack of data, and of the general optimism, to forecast blue sky scenarios, luring in buyers. Rising share prices then act as (false) evidence the forecasts are correct.

Typically, new companies are valued using new (i.e. unproven) key indicators; like clicks or eye balls or users, no matter if the latter are paying customers or not. Some even use costs as a key indicator, when there are no sales. The more costs, the more losses, the higher the value… Biotech analysts have been trail blazers within ‘cost valuation’. The more biotechs spend and lose, the more they’re worth -until they’re not.
 
 
Let’s not forget about the smoke and mirrors of three-letter abbreviations.
 
There are only so many three-letter combos to go around. Consequently the same three letter combos mean diferent things in different industries and at different times.
 
That’s not a flaw, it’s by construction. It’s supposed to create confusion and make the client or outsiders feel inferior, afraid to look stupid. The companies first make some analysts feel stupid who then turn around and fool their clients. Oh, it’s a WAP company! Buy. They have a BTC mechanism and make a ton on OEM…
 
Don’t fall for the simple three letter abbreviation trick
 
Ask and ask again until everything’s clear. Then make reasonable forecasts not based on 100% market share forever, constantly rising margins, and world domination of this industry and ten more nearby.
 
Take a minute to remember all those formerly glorious companies that perished and disappeared in the 2001-2003 downturn, despite all the thre-letter combos in the world in their Power Point presentations.
 

Don’t forget geographical trends, product trends… to see if, e.g., the first markets, the home markets are losing steam, or perhaps if new markets are small but growing quickly – which might not be visible in overall numbers otherwise. Slice and dice numbers in all directions and look for S-curves, tapering, bumps and exceptions.

Typically it takes at least a few years to get reasonably good at doing fundamental research, and at least a week to build a basic valuation model for a single company. It’s hard work and it’s complicated but it’s worth it if you have a nest egg worth a handful of years’ income to invest.

Professional analysts have spent years on their models, but that’s going overboard in my opinion and only meant to impress clients.
 
Equity research is at its foundation very simple; don’t let anybody tell you it’s best left to professionals. Just make sure you know what you know, and be honest to yourself about what you don’t know and what risk that entails
 
 
Less is often more
 
Most important of all, don’t make your model larger than what you’ll actually keep updated.
 
equity research
 
Don’t fill it with more data than you’ll actually use.
 
Erase whatever is unnecessary. Ask yourself, does this particular data series affect the final valuation in any way. If not, get rid of it.
 
And, again, it’s not true just because you wrote it down, or based it on an extrapolation of series of numbers. The future is inherently unknowable. It can’t be foretold. Less things happen than could have happened and we never know fully beforehand which one’s it’ll be. We don’t even know ex post what caused a certain future to come into being.
 
Don’t supersize that model, please
 
 
 
Stopped clock markets – or analysts

Stock prices and stock markets very seldom reflect the true long term value. Just as a stopped clock shows the right time twice a day (in Europe, actually only once), the stock market is valued more or less correctly about every six years or so.

That’s a good thing. It’s something you can profit from; if you have the guts to take a contrarian view when everybody else is screaming sell or buy. Or the guts to run with the lemmings for a while longer, even after the fat lady has finished her aria.

Markets are not weighing machines, they are beauty contests (in the short to medium run). They don’t find the correct value of a stock and stay there. On the contrary, markets forces of fear and greed create trends where the (perceived) prettiest stock is bought and the ugliest sold, until something, anything, changes and the trend reverses.

 
This makes stock prices and markets oscillate (in cycles of decades) around the true value. In the short term prices oscillate around some perceived popularity value.
 
The long term cyclicality creates twin opportunities for you as an investor: 1) follow the momentum, and 2) buy low, sell high. I think you should try to do it all:
 
  • Buy whenever a stock or an index is reasonably valued, or cheap, even if the trend is negative. Keep buying small increments.
  • Ride the momentum until it seems to stop (market dispersion crosses a threshold) or valuations venture too far out in the 5% tails
  • Sell if it is extremely expensive, no matter the trend, but don’t go short until market dispersion tells you risk aversion is high enough.
A precise depiction of professional investors in action
 
 
 

Hype and Hope is not a strategy
 
 

Don’t rely on blue sky scenarios. Very few companies hit it big. Why would yours? Perhaps just because you are looking at it (quantum mechanical interpretation).

 
Inoculate yourself against rosy forecasts by checking old stock market tables and company histories. Where just a handful succeeded, hundreds fell into oblivion or exhibited mediocre numbers.
 
Remember that you are only looking at the winners when looking at the current stock market
 
In fact, even the very best, the hundred top cash flow growers f all time, the Warren Buffets of listed companies, didn’t grow faster than 20% per annum for more than 20 years. OK, that fact needs to be checked, and many do manage to grow 100%+ per year for a few years, making compound numbers very impressive.
 
However, once growth falls to 20-40% per year there are rarely many years of >20% growth left. But, don’t take my word for it, look for yourself in the industry your current pet stock is operating in.
 
 
Status: It’s complicated
 
In the end it all is unbelievably complicated, despite there being just two possibilities at any given time: Up or Down?
 
There are thousands of professional investors, if not millions, trying to make money on the market, using different strategies. There are scores of companies competing for the same client wallet share. Remember that company management often get their forecasts completely wrong at turning points, so how could you beat them at their own game?
 
Don’t just flip a coin and hope for the best. Do the only thing very few investors do.
 
Pay attention to long term fundamentals, be patient and only buy when you have a decent safety margin, in terms of returns and think you can hold on forever and be pleased with that.
 
Don’t run with the crowd, unless you know what you are doing.
 
  • Don’t hope to sell the stock to a bigger fool
  • Don’t hope for efficient markets having priced all shares correctly so you can just wander in and buy blindly
  • Don’t hope that just any old stock will work as a hedge against inflation and money printing.
 
 
Equity research, explicit summary
 
  • Equity research is really easy and impossibly complicated at the same time. Just as life in general. Decide what type of effort you are willing to put in and what level of risk you accept. Just be explicit about it
  • Decide what kind of an investor you are. Active? Fundamental? Single stocks or index? Technical overlay? Gambler? Act accordingly
  • Do the math properly if you are going to do it at all. There is no big secret to forecasting; just do it, and be clear about your premises at all times. Make plausible and reasonable forecasts that are internally consistent. Double check your variables repeatedly. Believe the hype if you want; make a leap of faith in a product if you want, but make sure you understand that is what you are doing. Be explicit in your extreme assumptions.
  • Cheap or not? Compared to what? To your required return? To your marginal borrowing costs (your most expensive loan if you have any), to what you think the market’s marginal required return is or will be in the future? Compared to other stocks? Here is an overview, a check list, for doing equity research.
  • Decide on timing: It may be cheap, but markets are trending sharply downward. Then don’t buy until things calm down, or be prepared to ride a long downturn. Pace your purchases, buy a tenth at a time unless you have a specific game plan.
  • Wait. Be patient. Don’t expect good opportunities just because you are looking right now. It may take years for real opportunities to arise, whether it is about stocks, finding a life partner or a dream job. It can take 20 years for fundamentals to manifest themselves (though 8 years usually is enough). If you bought on the right side of cheap, time will work for you. In the meanwhile you accumulate dividends and perhaps buy more shares.
 
 
Addendum: The blue sky scenario is Amazon trading down by 90%

My best guess is that Amazon’s sales growth will fall slightly each year from 17% to 15%, from 15% to 10%, and sooner or later from 10% to 8 to 5% (assuming around 5% nominal growth rate for the economy in general).

Just spending a few hours with Amazon’s numbers, I think it is 10-20x as expensive as it should be. The current growth rate is only 17% per year, despite expanding its staff by 38%, and the company is hardly making any profits.

A year ago, when I last looked, sales growth was 20% and employee growth 36%. Fundamentals are not getting any better.

Assuming a decent 15% ROE in the future implies 1.5-2% operating margins and not even a 0.5% earnings yield (net profit/market value).

I would require at least 10-20x that return, i.e. 5-10% annual return on my investment, to bet on Amazon succeeding in its endeavor to sell everything to everybody. Consequently, the stock price needs to fall by 90-95% to satisfy my return requirements.

In addition, I can’t see any reason to assume growth or margins expand meaningfully in the future. Rather, I am worried growth will taper faster than most expect and that margins might fall to zero or below. I know the market believes and assumes Amazon can raise its margins to 5-10% as soon as it wants to, as soon as it concludes its hypergrowth phase (17%, remember?). It’s just that you seldom see those margins in retail. Ask Wal-mart.

In my opinion the blue sky scenario is Amazon trading down by 90%. The base-line and worst case scenarios are that Amazon falls even more and becomes completely irrelevant and taken off the market.

So, what should you do? Well, not buy Amazon, that’s for sure.

There are tens of thousands of other listed companies worldwide, many of which look more interesting than Amazon.
 
And if they are all too expensive, just wait. I’m sure you will get a good opportunity within the coming ten years, and in my opinion, most likely within two.
If that seems too long to you, you just aren’t fundamental enough. Go play with the chartists instead.
 
 
Final words; do this
 
1. Make the simplest possible models for your current investments. Spend about an hour (or less) on each company to begin with. What sales, margins, and valuations are required to make you happy? When? Is that plausible? If not, contemplate selling – not least given the extremely high general market valuation environment. But read my disclaimer first.
2. Sign up for my newsletter and free e-Book to stay up to date and learn more
3. Share this article with a friend that’s gone deep into story stocks