Why the Dow Jones Industrial Average is useless

The Dow Jones Industrial Average (or ‘the Dow’, ^DJI, DJIA), is a stock market index created back in 1896 by Charles Dow, a journalist. The goal of the index is provide a view of the stock market’s health. The Dow is an index of just 30 (arbitrarily chosen) stocks, weighted by the stock prices of its component companies and nothing else. To calculate, simply add up the 30 stock prices and divide the total by the Dow Divisor. When the index was formed, the divisor was 30, but after over 100 years of compenent changes, dividends, and stock splits, the Dow Divisor is around 0.14602128057775. The Dow is the most cited and most widely recognized of the stock market indices, and below I’m going to walk through why it’s completely meaningless.

At the highest level, the Dow is not the best represenation of stock market performance, simply due to the low (30) number of companies included in the index. However, in addition to only covering 30 companies, the Dow is weighted on company share price, not market capitalization (company valuation) as it should be.

The Dow is weighted on share price, not market capitalization

As mentioned above, the Dow simply adds up the share prices of its 30 companies, and divides by the Dow Divisor. This means that a given company can have an influence on the index disproportionate to its actual value. Market capitalization would be a better way to weight each company’s influence in the index’s movement, since it provides the actual value of a company traded on the stock market, calculated by multiplying share price by total outstanding shares.

To give an example, say a fictional company, ‘Corporate Synergies’, has a per share price of $100, but only 10 shares outstanding, meaning a market capitalization of $1,000. Next, lets say another fictional company, ‘Leveraging Innovation’ has a per share price of $10, but has 900 shares outstanding, meaning its market cap is $9,000. Now, in this example the weighting of these companies should be:

  • ‘Corporate Syngeries’: ($1,000 market cap)/($10,000 total market cap in index) = 10%
  • ‘Leveraging Innovation’: ($9,000 market cap)/($10,000 total market cap in index) = 90%

However, applying the price-weighted methodology used by the Dow, we arrive at basically the opposite weighting:

  • ‘Corporate Syngeries’: ($100 share price)/($110 total share price in index) = ~91%
  • ‘Leverage Innovation’: ($10 share price)/($110 total share price in index) = ~9%

So, as we can see above, since the Dow relies on share price weighting to move the index, lower value companies can be overrepresented, and higher value companies can be underrepresented — remember, market cap = company value.

Now that we’ve gotten through the basic argument along with a small example, here’s a chart showing the weighting of each company in the Dow by share price (incorrect, but used in Dow) and market cap (correct, not used). As you can see, there are in fact many grossly misweighted companies in the Dow today.

So, next time you hear someone talk about the Dow in any serious sense, kindly point them in the direction of an index weighted by market capitalization (e.g. S&P 500 or Wilshire 5000), that consists of more than 30 chosen companies. Together, we can stop making one of the most meaningless stock market indices the most widely cited.

If you’re interested in the specific companies in the Dow, and the associated weights, check out the table here.

Also, enter your email below if you want to be notified when I post once/month or so (I don’t spam, pinky promise)


Salary distribution, earnings per minute, and career earnings across professional sports leagues

Average/median salary by professional sports league:


In 2016, the average MLB player made nearly $11 million dollars per year in salary alone. This is more than an average NBA ($6.5M) and NFL ($4.2M) player combined, or the equivalent of 34 average MLS players.

The median salaries show us a similar trend, with each league’s annual salary shifting slightly lower than the average above; as shown in the distribution of salaries further below, there are a good deal of high earners that skew the average up a bit.


Salary distribution by league

Here, we can see that the top 25% of MLB players earn more than ~$16M per year (top whisker); this is more than any NHL or MLS player, and more than 99% of EPL, 97% of NFL, and 90% of NBA players.




Earnings per minute played

Below, I split out earnings per minute of regular season game time, both excluding and including stopped clock/interruption time. Here, we can see that NFL players top the list, earning $2,709 per minute of running game time, and $847 per minute of total game time (includes timeouts, halftime, etc). Interestingly, the top earning league MLB falls to 5th on our earnings per minute list due to its 162 regular season games.


Another interesting view that I don’t have the data for would be to calculate earnings per minute of total work, including practice/training, press, travel, etc. I imagine this would trend similar to our average annual salary list above, with most athletes putting in 8+ hour days irrespective of how many games they have in a season.

Earnings over career

The average professional sports career is by no means a long one, with many exiting the competitive environment due to injury or declining performance. So, when comparing earnings across leagues it’s also worthwhile to factor in the average career length.


Here, we see the divide between MLB players and the rest of the groups grow even wider, as their annual earnings are multiplied by their lengthy careers (only second to the EPL).


  • Unsurprisingly, the top few earners in each league make orders of magnitude more than the rest
  • MLB players earn the most ($11M/year), and have the lowest earnings standard deviation (e.g. a lot of players earn a lot of money)
    • The median MLB player earns nearly $50M over his career, while the average player earns even more
  • NFL players lead in per minute earnings, earning $2700 per game minute (regular season only)
  • MLS players do not make much, relative to the other leagues shown here
    • You can pay 1 ‘average’ MLB player, or use that money to start an MLS team comprised of 34 ‘average’ players

If you enjoyed this post, be sure to join our email list below to be notified of new posts every sometimes. (I don’t spam, share your email, etc)

What type of companies does Y Combinator fund? [YC exploratory analysis]

A bit of background on Y Combinator, for those unfamiliar

Y Combinator (YC) is an American startup incubator founded in 2005 that has invested in many well-known companies, such as: Dropbox, Airbnb, Coinbase, Stripe, Reddit, Zenefits, Instacart, and Twitch.tv. Its portfolio of companies have a combined market capitalization of ~$65B [though, it’s worth noting that many of the companies are still privately held/reliant on VC funding, meaning the valuations are a bit loose].

YC invests and incubates in companies bi-annually, with summer and winter groups. YC receives applications from founders looking for seed funding, help developing their product or service, and help scaling their business. Of these applicants, a select few are accepted into the program. The data in this post comes directly from YC’s website and covers companies funded by YC from its inception to October 2016.

YC has funded >880 companies since 2005, and has generally been increasing the number of companies it accepts each year


No surprise here that Y Combinator has been increasing its number of investments per year, as it begins to realize success from early plays and the notoriety of the firm grows.

YC funds winter and summer classes of companies each year, below are the seasonal trends:


YC has funded 450 Summer class companies and 433 Winter class companies, with similar batch sizes by year for both.

YC investments by vertical has invested in 10 major verticals, with B2B and Consumer topping the list


YC investments by vertical by year


As we can see, YC still is liking B2B and Consumer companies, but have significantly increased investments in Enterprise, Hardware, and Biomedical companies over the last 2-3 years. (An interesting additional view for this would be to normalize the sector n-counts for sector size, bonus points if someone does this.)

Most common words in YC company descriptions


I ran text frequency analysis on the YC company descriptions, stripped out most of the junk words, and was left with this list. This isn’t a super telling chart, but I did find some of these a bit interesting, so I dropped it in here. Looks like you should start an Enterprise, Hardware, or Biomedical platform company if you’re looking to get accepted into next season’s program. 🙂

The data/iPython notebook

As mentioned, this list was scraped from Y Combinator’s website, and I used Python to make the charts. If you’re interested in checking out the iPython notebook, it can be found here.

Subscribe to our mailing list to be notified of new posts + receive occasional extras

* indicates required

Understanding the price-earnings (P/E) ratio, and a few of its limitations

The price-earnings ratio (P/E Ratio) is a ratio used to value a company that measures its current share price relative to its per-share earnings (profit per share). The ratio can be calculated as:

Market Value per Share (Share Price) / Earnings per Share (EPS)

Essentially, the P/E ratio is meant to show how much investors are willing to pay per dollar of earnings. In general, a high P/E suggests that investors are expecting higher earnings growth in the future compared to companies with a lower P/E.

The first question many have when seeing the above formula is “what happens to the P/E ratio when a company has negative earnings?”. This is an excellent question! Although companies with negative earnings can technically have negative P/E ratios, the standard practice is to display ‘N/A ‘, or show no ratio for that quarter’s P/E. However, a single quarter of negative earnings can skew the P/E for the next 3 quarters (when earnings are later positive and P/E is again displayed), since the metric adds the trailing 4 quarters of earnings; this means that a negative quarter can make the denominator (EPS) of the above equation very small, sending the P/E ratio through the roof. I’ll highlight a couple current examples below when I split out the companies with the highest P/E ratios, which may make this point easier to follow.

P/E ratio by sector (averages can be deceiving)

The below chart shows us the various average P/E ratios across sectors, for the largest ~500 publicly traded companies. At first glance, looks like Materials and Health Care companies have sky high P/E ratios:


However, if we bring in the standard deviation of the P/E values in each sector, we see that Materials and Health Care both have very high standard deviations, meaning some companies likely have unreasonably high P/E ratios (I’ll get into the cause of some of these outliers below).



So, since our averages are skewed due to outliers, the median P/E would probably be a better metric to use:


As we can see, P/E varies pretty significantly between the different sectors, with Real Estate, Energy, and Health Care topping the list. It’s worth noting that P/E ratios can vary across sectors due to differing ways in which companies earn money, and differing timelines in which that money is earned; as a result, P/E is best used as a comparative tool when considering investing in companies within the same sector.

That said, I did pull the top 10 companies with the largest P/Es below, in an attempt to highlight some of the reasons a company’s P/E can be several standard deviations above its sector average.

Which companies trade at the highest P/E ratios? (understanding unreasonably high P/E ratios)


At first glance, one may look at Glaxosmithkline or Rio Tinto and think investors must either be crazy or expect a tremendous amount of future earnings growth, since the companies are trading at 1.2-3k multiples of earnings.  However, a deeper look shows that Glaxosmithkline had negative net income in 2 of the last 4 quarters, while both Rio Tinto and Kinder Morgan also had a negative net income quarter at some point in the last year. (note this is the example I mentioned I’d cover at the beginning of this post)

Basically, companies with sky high P/E fall into these basic groups:

  1.  The company had some negative earnings in the trailing four quarters, but not in the most recent quarter (if the earnings were negative in the most recent quarter, the general practice is for no P/E metric to be shown)
  2. The company has been squeaking out very small profit margins (and consequently a small EPS), while their share price is reasonably high
  3. The company is actually expected to grow its earnings by a large amount so investors are happy to buy large multiples of current earnings (this is the assumed case with Netfllix, Salesforce, and Amazon)
  4. That’s not all! Debt, leverage, earnings accuracy, one-time purchases, and more can impact P/E ratios as well

The takeaway from this is that P/E ratios should definitely be paired with other metrics to get the complete picture. Many financial metrics can skew for a variety of reasons, and this is helps illustrate how important it is to avoid simply looking at a single financial metric and pulling the trigger on an investment. 

Also, for those interested I pulled the 10 companies trading at the lowest P/E multiples (note they mostly fall into sectors that bottomed out our list above)

Which companies trade at the lowest earnings multiples? (are expected to grow the least) 



The data

I used Fidelity’s stock screener to grab the data, though in hindsight would be easier to pull directly from Yahoo Finance. I then used Python to clean with Plot.ly to make the charts. Those interested in the iPython notebook can view it here.


How much free time do you have?

Since I graduated college a couple of years ago, I can’t really remember a time where I’ve been ‘bored’ for an extended period, or felt like I needed more to do. If anything, since I left college I’ve felt strapped for time, weeks blurring into months without accomplishing much outside of working and taking care of myself; this is a bit off-putting, as I don’t have kids (or pets), I have a good deal of energy, and I try to be smart about my time.

Looking to quantify where the 168 hours in each week go, I created a small calculator. Basically, this is designed to identify how many hours of free time one has in a given week, based on the selected inputs. Additionally, since time allocation drastically changes post-retirement, this calculator asks you to pick a retirement age and will calculate your free time until retirement.

Free time calculator–calculate your free hours per week, and until retirement [or death if those are synonymous!]

Note that this is free time for a typical week, so excludes both vacation days and one-off ‘busy items’ (like taxes, car/house maintenance, etc), but should give a good approximation of free time available.

For my particular case (defaults in calculator), I have about 28/168 hours free in a given week, or 17% of my week. Additionally, I’ll have the equivalent of ~6/37 free years from now until my lofty retirement plug of 60 years old.  This is actually a bit more free time than I was expecting, but I speculate I’m either underestimating something, or I waste a decent amount of time on the front/back-end of various activities, as there’s a transition time into some of them.

Inputs are all adjustable (and should work on mobile with some scrolling), so free to use the calculator to determine your own amount of free time available!

Which industries correlate closest with the S&P 500?

I used python to calculate the correlation between stock returns  listed in the S&P 500 over the past ~12 months (more notes on method at bottom of the post). I cut this data a couple of ways, which I’ll cover below:

  1. Grouped stocks by industry, and calculated which industries correlate closest with the S&P 500 index (I used the SPY, an ETF designed to track S&P 500 movement, as my proxy)
  2. Grouped stocks by industry, and calculated how correlated individual stocks are within a given industry (said another way – how related are the returns of a set of stocks in each industry?)

Pearson’s correlation coefficient

(tl;dr: in the charts below, the higher the # the more correlated the variables are)

I used Pearson’s correlation coefficient for the purposes of this analysis; it’s a measure of linear correlation between two variables, and can be used to see how tightly correlated different stocks are. The output of the value must be between -1 and 1, and is defined as such:

  • Positive values denote positive linear correlation
  • Negative values denote negative linear correlation
  • A value of 0 denotes no linear correlation
  • The closer the value is to 1 or –1, the stronger the linear correlation

Which industry correlates closest with the overall movement of the S&P 500?

Screen Shot 2016-08-23 at 11.42.20 AM

We can see here that stocks making up the Industrials sector within the S&P 500 track closest to its overall movement while Consumer Discretionary has the weakest correlation with S&P 500 movement.

There’s no relationship between total industry market cap and the S&P 500.

Astute observers may note that the S&P 500 index is calculated as the weighted average of the market capitalization of the 500 equities it’s comprised of–meaning industries with the largest total market caps would, in theory, have the largest influence on the S&P 500’s movement.

Interestingly, this is not the case:

Screen Shot 2016-08-23 at 12.01.11 PM

The r-squared value is 0.09, indicating there is no significant relationship between the total market cap of an industry within the S&P 500 and the SPY’s movement over our time frame of 1 year.

How correlated are individual stocks within a given industry? (or, how volatile are the returns in an industry across its equities)

Next, I calculated how tightly correlated individual stocks were within a given industry. Below, we can see that Utilities tops the list, followed by Telecom.


This means that, if I have a portfolio comprised solely Utility stocks, I don’t have much diversification since the different companies operating in that sector tend to move in lockstep. Additionally, it means that I can have a pretty well-diversified portfolio comprised of equities solely in Consumer Discretionary, or Healthcare, since those industries have weakly correlated equities.

Some notes on this analysis:

  • Correlation does not imply causation.
  • Time frame covered was 8/10/2015 to 8/10/2016, or roughly the past year’s worth of stock movement data. Different time windows will yield different results, meaning these measures are only accurate for the time window specified.
  • I am comparing the correlation of the stock’s returns using the month-end close price from Yahoo (or Verizon :p) data. This does not account for dividends, but it seems to be a decent approximation 
  • This isn’t a particularly refined way to do this (there are people who do this for a living), but it acts as a good approximation
  • Disadvantages of using Pearson’s correlation coefficient:
    • The approach is only valid for linear dependencies, which are not always observed.
    • The approach only captures the first two moments of the relationship. This means a value of 0 does not necessarily indicate a relationship does not exist.

Walking is 3 times more expensive than driving, from a fuel perspective

Edit: I am not advocating people stop exercising or drive everywhere, this was just a simple stat limited to the scope mentioned below. There are of course health benefits to walking and environmental trade offs associated with driving. 

Inspired by a Wikipedia article I came across on energy efficiency in transportation, I decided to compare the cost of the various transportation methods. In order to do this, I split the cost calculations into the following two groups:

  1. Self-powered transportation (walking, running, and biking): I calculated the average calories burned per mile from the mentioned self-powered transit modes, along with the associated cost per calorie. Using this information, I could then determine how much it would cost per mile to “fuel” this transportation.
    • Fuel cost per mile = calories burned per mile/cost per calorie
  2. Motor transportation (cars, trains, etc): First, I pulled the energy (in kWh) required to travel 1 mile for each mode from above Wikipedia article, along with the gasoline equivalent of said energy consumption.  Next, I paired this with average gas costs to determine the energy cost per mile.
    • Fuel cost per mile = gasoline required per mile/cost of gasoline

note: there is also some basal metabolic rate each of us has that will burn calories during non-active motor transport, but the assumption here is that the caloric burn through walking, biking, and running are additive to this base rate, meaning it’s ‘extra’ fuel required

Fueling walking costs 3 times more than fueling your car:

Surprisingly, from a pure energy perspective (using the methodology mentioned above), biking, walking, and running are the three most expensive types of transportation listed:

Screen Shot 2016-08-07 at 10.16.43 PM


Now, the above chart uses the average American daily food expense of $7.00, paired with an average caloric intake of 2500 calories per day, for a $0.0028/calorie cost. Some of you may look at this $7.00 per day and say, “I can do better than that”; to this I say, how about a McDonald’s-only “complete/nutritious” diet running you $4.40 per day or $0.00176 per calorie (a ~37% reduction from the American mean).

Sadly, biking, walking, and running still check in as the most expensive forms of transportation to fuel.

Even on a McDonald’s only diet it’s more expensive to fuel walking compared to driving:

Screen Shot 2016-08-07 at 10.17.41 PM


I know, I know, McDonald’s isn’t that cheap. We can do better. How about only eating (drinking) pure vegetable oil (aka pure fat) to fuel these activities? A gallon of vegetable oil will run you $5.76 and provide 30,720 calories, for a cost per calorie of $0.000194 (a 93% reduction in food expenses from the American average, and an 89% reduction from our McDonald’s scenario).

A vegetable oil only diet will make biking, walking, and running cheaper to fuel than driving:

Screen Shot 2016-08-07 at 10.18.17 PM


Considering total cost of car ownership in calculations:

The above scenarios were simply comparing fuel costs between the transportation types, but there are, of course, a number of additional costs tacked on to owning a car. Using AAA’s average cost per mile of ownership of $0.34, and pairing it with our already-calculated fuel cost we arrive at a $0.40 per mile total car transportation cost.

Plugging in the total cost to own for the car, and comparing with biking, walking, and running, we can see it’s anywhere from 33-400% more expensive to drive, depending on activity type.

Screen Shot 2016-08-07 at 10.21.42 PM

*note: I’m not layering any additional costs onto biking/walking/running so if you have a bike that costs more than some cars, or buy new shoes every week then these numbers will increase

Why private companies are delaying IPOs, and why we need a new definition for ‘Technology’

The well-known and highest valued Tech companies are delaying IPOs (think Uber, AirBnB, etc)

According to McKinsey research, at the end of 2015, 146 private Tech companies were valued at ‘unicorn’ status–(a private valuation of $1 billion or more). Clearly, the private market for Tech companies is skyrocketing; in fact, Tech investors laid down $76 billion in investments in 2015 vs. $26 billion in 2013.

There are many reasons as to why the private market is growing, valuing more and more companies upwards of $1 billion through late-stage investment rounds (justifiably or not). However, I’d like to focus on why so many companies are opting to delay going public, and why they will eventually IPO.

Why these companies are staying private:

  1. The first, and most obvious reason, is that U.S. law changed to increase the max # of shareholders allowed for a private entity. Previously, just 500 shareholders were allowed before an entity needed to go public, now a business can have up to 2000 investors. Companies like Facebook and Google were essentially forced to IPO as they triggered the 500 shareholder limit; that limit has just quadrupled, greatly extending the window companies have to operating privately.
  2. Second, staying private offers the advantage of being able to focus on long-term strategy, rather than short term quarterly earnings reports. Once a company IPOs, earnings growth is expected, and improvements each quarter into perpetuity are demanded from Wall Street investors; you can see how this would not blend well with some of these large private Tech ‘startups’, who are much more focused on generating user growth and network effects for the time being, opting to forgo large chunks of profits until they reach desired scale.
  3. Third, most private Tech companies offer their employees a combination of cash (salary) and equity-based compensation. The employees then can’t liquidate this equity until the company goes public (though some do allow private market share liquidation), meaning they are essentially tied to the company until it IPOs; this can be an effective strategy to reduce attrition and keep talent in-house much longer than industry averages.

Why they will eventually move public:

  1. As mentioned,  the max # of shareholders an entity can have before moving public has increased to 2000. However, it’s still a max, and will no doubt trigger a lot of the private unicorns into going public.
  2. Although a lack of liquidity can be beneficial short-term to keep talent in-house, having liquidity will eventually be desired as some of the company leadership wants to cash out, and will be necessary to scale the company up and bring in new talent as the market catches onto this private equity trap.
  3. Going public will allow for easier financing, allowing the now-private companies to scale their businesses even larger with much less difficulty. Take Elon/Tesla, for example, he’s able to make significant capital investments, and even acquisitions (SolarCity), using his public market cap.  Elon has been able to swing up market cap based on forward-looking statements and pre-orders, which has been an extremely powerful tool for his company and is core to Tesla’s future success.

Although the ‘unicorns’ are staying private, worth noting that the number of Tech IPOs is not decreasing

Lately, there has been plenty of news focusing on the lack of Tech IPOs. I decided to dig into this, grabbing IPO data from the NASDAQ and NYSE, and found that the # of Tech IPOs is not as low as some make it out to be. While it’s certainly true that many of the more famous ‘unicorns’ (Uber, Pinterest, Palantir, etc) have been delaying their IPOs, I think it’s important to note that the # of Tech sector IPOs has still been quite high, hovering between 20-40 per year over the past 5 years.

Screen Shot 2016-07-29 at 9.29.28 PM

In the above chart, note that as a % of total IPOs, Tech sector IPOs are indeed decreasing; I will address this towards the end of this post.

The definition of Technology needs to go:

The below heat map shows the # of IPOs per year, by sector:

Screen Shot 2016-07-29 at 9.55.41 PM


Cutting the data to show the last ~15 years:

Screen Shot 2016-07-29 at 10.02.37 PM


From the chart, we can clearly see that both Healthcare and Finance have had a ton of IPOs in the last ~3 years.  Digging into the growth in Healthcare IPOs, specifically, I noticed something interesting: technology was the driving influence of many (>30%) of the Healthcare IPOs. Of the list of Healthcare IPOs, nearly a third were coming from either Biotechnology, Healthcare Electronics, or Healthcare Instrument companies. It can be argued that all three of these industries utilize and rely on technology to a high degree.

[for those interested in deeper dives into Healthcare, check out my last two posts: (1) Healthcare expected to grow faster than any other industry through 2024 (2) Two forces driving Healthcare’s growth]

What exactly is Technology, anyways?

Technology is defined as

“The application of scientific knowledge for practical purposes, especially in industry.”

Using this definition, it becomes apparent that companies scattered across nearly every sector could quality as “Tech” companies. Perhaps although the dictionary definition is pretty open-ended, the definition of the Technology sector as defined by the markets is more complete. Taking the definition of Technology sector from Investopedia:

“The technology sector is the category of stocks relating to the research, development and/or distribution of technologically based goods and services. This sector contains businesses revolving around the manufacturing of electronics, creation of software, computers or products and services relating to information technology.”

Blurred lines:

From the above definitions, we can see the line really beginning to blur regarding what’s a Technology company and what’s not. Take a company like Uber, for example–Uber provides logistics/transportation services directly to consumers, but considers itself a Technology company since its business has been enabled by technology, and it’s a critical component of Uber’s future success. Another example, Amazon, essentially began as an online bookstore and eventually transitioned into a full-fledged online retailer known for its world-class logistics (ignoring their recent Cloud offerings). Rather than initially being labeled as a publisher or retailer, Amazon has been deemed a Technology company from the start, since their business was enabled by technology.

Technology has already drastically changed Retail and Transportation (among other industries), and is quickly becoming a core component of every company’s strategy, agnostic of industry. Every major company will eventually need to adopt the latest technology in order to stay competitive over the long-term. This brings us to my core problem–if every company needs to adopt technology in some way, and the definition of a ‘Technology company’ is already so loose, one can eventually argue that every single company is a Technology company, rendering the label useless. Finance has already run into this to some degree, creating a hybrid industry, FinTech, to describe companies that ‘use technology to make financial services more efficient’; I’d argue that every company in the Financial sector is, to some degree, attempting to use technology to improve efficiency. Are we going to start calling Oil and Gas companies who leverage technology to find new drill sites OilGasandTech companies? Or are Consumer Products businesses who adopt the latest automation technology into their factories going to be named ConsumerProductsTech entities? Where does it end?

Clearly, there needs to be some refining of the definition of what constitutes a Technology company and what doesn’t. I don’t have a perfect answer, and perhaps a new term will present itself eventually. For now, I think Technology companies should be limited to those who are solely in the business of creating and selling technology, while all of the businesses who leverage technology to gain a differentiated advantage in a given industry should remain, well, in the industry they are attacking.

Two major forces driving the rapid growth (+7 million jobs by 2024) of Healthcare in the United States

I recently published a post detailing industry job shifts in the United States, and one of the key takeaways was that healthcare is expected to grow faster than any other industry, adding ~6.8 million jobs in the next ~10 years. Politics aside (looking at you, Obamacare), below are two reasons healthcare is expected to add so many jobs in the next ~10 years:

*note, I do not claim that these forces are the sole factors that will drive healthcare growth, just major contributors

(1) The population in the United States is aging…fast:

Below, we can see that as a percent of total population the 65-84 year-old cohort is expected to grow by 4% in the next 10 years.

Screen Shot 2016-07-16 at 10.26.11 AM


What does this translate to in terms of actual people, you may ask? The 65 and over cohort is expected to grow from 47.5 million to 65 million (+17.5 million) in the next 10 years.

Screen Shot 2016-07-16 at 10.27.28 AM


(2) Life expectancy in the U.S. is increasing by about 2 months each year, moving from 70 in 1960 to 79 years in 2014. 

Screen Shot 2016-07-14 at 8.55.19 PM

(Source: Worldbank)

Putting things together (with a crappy drawing!)

Generally, the 65 and over age group utilizes a greater amount of healthcare (particularly nursing homes, outpatient clinics, etc), meaning as this group grows (point 1 above), demand for healthcare should grow as well.

So, when we have a large cohort aging and entering the 65-84 year age group (what I have coined below as the ‘window of medical need’), and we couple that with life expectancy increasing by ~2 months/year (point 2 above), we get something like this:




We can see as our 75 million Baby Boomers begin to retire, they enter the ‘window of medical need’, with 17.5 million entering the 65 and up age group by 2025. Additionally, those already in the 65+ club are enjoying longer life expectancies. This means the ‘window of medical need’ (green bar) is gaining a large amount of people, while expanding/retaining those in the window for a longer period; it is these two factors that are major contributors to the rising demand for healthcare. In turn, healthcare has seen a huge spike in IPOs over the past three years.

By 2024 Manufacturing expected to lose 1.5MM jobs while Healthcare to add 6.8MM (U.S. industries at a glance)

Current state of industries in the United States

The Bureau of Labor Statistics (BLS) defines ~150 top-level industries in the United States economy, and the below 10 largest industries account for ~80% of the economy.


Industries expected to add the most jobs by 2024 (healthcare expands)

Overall, the United States is expected to add ~43.5 million jobs by 2024. The following 10 industries adding the most jobs are expected to drive >30% of that growth:



Next, these 10 industries are expected to have the largest % growth in jobs, with the highest, outpatient care centers, forecasted to grow by nearly 50% over the next 10 years!


Notice that healthcare dominates the list of high-growth industries; this is primarily due to two reasons:

  1. The massive ‘baby boomer’ population is beginning to age, driving up the total # of new individuals in need of healthcare
  2. Americans in general are living longer, meaning that individuals who are already utilizing these healthcare facilities will continue to due so for a longer amount of time. Every year added to life expectancy has a tremendous impact on total demand when there are millions of folks in each cohort of the healthcare population.

Industries expected to lose the most jobs by 2024 (manufacturing contracts)

Topping BLS’ list of shrinking industries in terms of total jobs lost is manufacturing, followed by governmental functions:

When we split the data to show expected % change in # of jobs, apparel/leather manufacturing, tops the list, with an over 45% reduction expected!


(not) Made in America/made with robots 🙂

Notice that the manufacturing industry is expected to contract the hardest in the U.S. by 2024. Overall, manufacturing of all kinds is expected to lose ~1.5MM jobs in the next 10 years, or 6% of its total workforce; the industry currently accounts for ~16% of the total workforce, and after this change it would drop to ~14%. This drop is due to increased automation capabilities and an increase in outsourcing manufacturing outside of the U.S.


Speaking of automation (looking at you, transportation industry)….

While on the topic of automation, I found it interesting that transportation is still expected to add ~255 thousand jobs into 2024, for a total employment of ~7.4 million (nearly 2% of the total working population).

While the true timeline for self-driving cars/trucks/etc. is certainly debatable, it’s tough to argue that this industry won’t eventually face a massive step-change/contraction in labor due to automation. —this is similar to what’s currently going on in pockets of manufacturing, as automation solutions continue to drop in cost, while labor costs continue to increase

The economic impact of a change in this industry could be severe, with the average truck driver earning ~$21/hr, and working ~41 hours/week, that gets us to an annualized salary of ~$44,000, meaning over $62 billion of salary could be impacted (coming from ~123M separate establishments) in the truck transportation industry alone. Extrapolating the truck driver’s salary across the rest of the above transportation industry yields over $325 billion worth of wages that could subject to change in the coming years. For reference, wages in the United States as a whole come in near $14 trillion.

Where will the $325 billion currently being directed towards wages be reallocated, if automation truly does eliminate the need for the majority of these jobs? A number of possibilities:

  • For starters, a good deal will be plowed into high-cost upfront investments into the automation technology itself for the first few years
  • Second, once efficiency gains are realized and OpEx begins to drive downwards, the value created will move to the shareholders, investors, and employees of the companies both providing and benefiting from the technology

In regards to the employees operating in the transportation sector, it’s of course unclear where they would move to. They could of course re-skill and move to other industries, but many other areas are also ripe for automation, meaning things could get pretty crazy. I’ve listed out a few options below of how things may shake out regarding automation of work:

  1. Automation doesn’t work as imagined, majority of current jobs are still needed to supervise the tech., not much changes
  2. Old industries are automated into obsolescence in terms of job prospects, but this  opens up the door for new industries/jobs we’ve never ever posited, and employees of impacted sectors shift towards these industries (see cars replacing horses, industrial age, Luddites)
  3. Automation fully eliminates the need for the vast amount of jobs/work we currently have today, no new jobs or industries are created as a by-product. Some sort of social safety net is put in place (e.g., basic income), our new robot friends perform all of the work, and we humans pursue leisurely activities, creative ventures, or higher-purpose undertakings (interplanetary colonization, immortality, etc)
  4. Same as option 2, but no social safety net put in place, a ruling class is created, people get pissed and overthrow them, does not end well for folks on top
  5. Same as option 2, but no social safety net put in place. A ruling class is created where a select few make a ****-ton of money and rest of the world faces low wages and high unemployment

Though the timeline is debatable, it’s clear that automation/technology improvements and outsourcing/globalization have already had a profound impact on the U.S. labor market, and will continue to do so.