Tuesday, March 8, 2016

Housing: Part 123 - Profit Margins and Efficiency

There was a recent dustup about a Yelp employee who posted an online essay complaining about how impossible it was to live in San Francisco on her very low wages.  Yelp fired her and moved some of its operations to Arizona.  I have been thinking about this issue of high costs in Closed Access cities, and this essay ties into the questions I have been wondering about.

I think Closed Access housing policies create economic rents that are shared by firms and workers.  Since the source of those rents is access to real estate, profits generally flow to real estate owners through real estate rents.  For low wage workers, since a large portion of their income goes to rent and they aren't able to make significant discretionary cuts to their housing budget, most or all of the rents that flow through them go to their landlords.

For higher wage workers, they can capture some of those rents by cutting back on housing expenditures.  And, since shelter is usually a fairly small expense for firms, they can capture much of the rents that flow through the firm.

Given this pattern, why isn't there more pressure to outsource non-skilled work to other locations?  There has been a strong pattern of high income households moving into the Closed Access cities and low income households moving out.  Why had this process not reached the low skilled Yelp position in the essay?

In fact, I think one of the peculiar factors that triggered the housing bust was that California has been systematically segregating by income.  The coastal core cities have been taking in high income households.  This is why we see activists complaining about gentrification or demanding building moratoriums.  This massive pattern of migration by income is the forest, and occasionally the newspapers find someone chaining themselves to a tree.

Migration Patterns Susceptible to Crises

Low income households, in an attempt to remain tethered to the lucrative labor market that Closed Access enables, move to the suburbs and exurbs.  This reverses the normal pattern of housing stock expansion.  As a society becomes wealthier, normally high income households improve the quality of the housing stock by building new, higher value homes.  They leave the existing housing stock to be claimed by lower income households.  So, in many cities, lower middle class housing is usually housing that was formerly upper middle class housing.  The 2,000 sq. ft. ranch style house that the doctor or lawyer built in 1970 with the latest styles of laminate countertops and vinyl floors becomes the aging 2,000 sq. ft. starter home for the construction worker who just got promoted to assistant foreman in 2005.

But, Closed Access means that location trumps everything, so now, instead of building new neighborhoods, high income households are going in to the existing cities and buying up the existing housing units in good locations.  This leaves the new neighborhoods to be built for lower income households.  In California, this describes whole cities in the Inland Empire.  So, in 2005, we had entire cities that were peopled by households with middle incomes, low net worths, and highly leveraged real estate.  These are the households most susceptible to an economic disruption.  But, because of Closed Access housing policies, now we had whole cities that met this description.

This is why we had the peculiar behavior of home prices during the boom, where prices of existing homes rose faster than prices of new homes.  To the conventional narrative, this looks like it could be explained by the influx of low income homebuyers, enabled by predatory lenders.  IW readers know that there was not an influx of low income homebuyers.  This is caused by the strange upturning of the norm in housing markets.  High income buyers were buying existing homes in prime locations, and lower income buyers were buying new homes in less valuable locations.  There was not a shift in the marginal buyer.  There was just a shift within the population of buyers about who were buying new and who were buying existing homes.

Normally, highly leveraged homes would be scattered among the existing housing stock as moderate income households moved into aging neighborhoods.  But, now, all these households were in one place, and when the disruption hit in 2006 and 2007, whole cities were devastated.  And, when it had become nearly impossible for moderate income households to obtain mortgages, the wave of defaults and devaluations moved first through these cities.

The New Economy and Profit Margins

Back to the main point, here.  It is interesting that Yelp would have positions worth $14/hour, and that they would choose to locate those jobs in the single worst place to try to live on $14/hour.  It seems clear that a position worth $14 is not a position that is capturing much value by being located in Silicon Valley.  Of course, there are many tertiary positions that are tethered to higher valued positions.  For instance, the janitorial staff for the executive suite must be staffed locally, and therefore funded based on the local cost of living.

But, I would expect there to be extreme pressure to move productive positions to other locations when the cost difference is so high and when labor is such a significant part of the firm's cost structure.

I think there is a complex web of causally dense factors here that form the messy foundation of our current economic challenges.  One of the defining characteristics of new economy firms in sectors like tech. and finance is that there are substantial reputational and network effects that lead to a sort of winner-take-all life-cycle for firms and competitors.  In tech., especially, lifecycles can be short.  This leads to two factors that increase profit margins.

1) Any firm can quickly and unexpectedly be knocked from their competitive perch.  This means that to justify investment in disruptive hardware or software, firms need to be able to earn back invested capital quickly.

2) Many potential competitors will not survive the early fight to be the leader in the category, so there is tremendous survivorship bias in the financials of existing firms.

So, we begin with highly risky investments that demand a higher return on investment, those investments are highly dependent on successful early-phase investment and implementation, which means that most of the costs of the eventual leaders will be sunk costs with very high gross profits, and by the time firms are recording profits, most of the sunk costs of the sector as a whole will have been written off the books of failed competitors.  So, we have a set of firms where the firm-specific profit margins are very high, but the sector-wide profit margins over the full life cycle of the sector are bid down to normal levels by competitive pressures, or are slightly higher than normal because of the competitive risks.

This means that there are a lot of firms like Yelp, who, when they pay low-skilled workers $14/hour, are vulnerable to rhetoric like the rhetoric in the linked article:
After posting a heartbreaking open letter to Yelp's CEO Jeremy Stoppelman about how impossible it was to live on her paltry wages, a Yelp employee named Talia Jane was fired. Yelp made $32.7 million last quarter. Jeremy Stoppelman is worth hundreds of millions of dollars. The post, on Medium, narrated the difficulties and indignities of daily life in one of America's most expensive cities for the people not blessed with stock options who are powering the tech explosion. Jane worked in customer service at Yelp and Eat24, a food delivery company that Yelp purchased in 2015 for $134 million:
"I got paid yesterday ($733.24, bi-weekly) but I have to save as much of that as possible to pay my rent ($1245) for my apartment that’s 30 miles away from work because it was the cheapest place I could find that had access to the train, which costs me $5.65 one way to get to work. That’s $11.30 a day, by the way. I make $8.15 an hour after taxes. I also have to pay my gas and electric bill. Last month it was $120."
San Francisco's minimum wage is $12.25 an hour, pre-tax, so it appears that Jane is making just slightly more than that. In a statement, Yelp said it won't comment on why she was fired (but insists it had nothing to do with the letter), but does agree that it's expensive to live in San Francisco—though doesn't acknowledge that it may be partly responsible for that state of affairs.
Later, Yelp's response on this problem is derided as "free-market dogma".

Now, if Yelp was running a local landscaping business, or a pizza delivery shop, they wouldn't have to worry about these accusations, because there simply wouldn't be any cash to spend.  But, here, since Yelp is the steward of a whole lot of value at emanates from their victories in those early phases of development, they look stingy.

But, why wasn't this job already located in Arizona?  I think this is part of a whole series of patterns set up by the Closed Access housing policies that now define our major cities.  Isn't it interesting that these frontier industries seem drawn to cities that can't house them?  I think this is a very complicated issue, and causality is messy.  But, could this be some sort of emergent outcome that arises from the extreme competitive pressures these companies face?

Let's say you've "won" your category.  Now, you have an outsized valuation and outsized profit margins.  If you were a car company 50 years ago, your worry, on a year to year basis was whether you might gain or lose 2% market share.  But, if you are Facebook, your worry, on a year to year basis, is whether you become Myspace or Friendster.  So, what if locating in Silicon Valley greatly increases your costs (which are very low on a cash basis) but decreases your chance of becoming Friendster by 10% every year.  That's a no brainer.  You pick up your operations and you move them to Silicon Valley.

This brings us to Yelp and Ms. Jane.  Is it worth having her position in Silicon Valley?  Well, compared to Yelp's $32.7 million in quarterly profits, Ms. Jane's $14/hour is paltry.  But, what if the ease of communication supported by having the operation she works in located in Silicon Valley decreases the existential risk for Yelp by 0.001%.  Probably worth the cost.  Now that Yelp is the category winner, those profits are well-used by building an expensive moat around the organization.

Partly, I think there are inefficiencies here, which in finance have been well-covered, regarding firms with large cash holdings or high profit margins.  When the risks to a firm of inefficient operations or financial management aren't imminent or palpable, it is difficult for the discipline of the marketplace to work sharply.  So, this state of affairs is a combination, I think, of inefficiencies and of complex efficiencies that tend to be misunderstood.

But, I'm afraid, in either case, Ms. Jane's position probably should have been located in Arizona to begin with, and the commonly prescribed solution to her problem, to expand policies like rent controls in a city already toppling under the weight of such incongruities is to simply double down on the Closed Access policy framework at the heart of the problem.

But, maybe firms like Yelp would never have been funded in the first place if that moat didn't exist.  Have Closed Access cities become valuable because frontier investments have taken on a winner-take-all character?  Or have frontier investments taken on a winner-take-all character because Closed Access cities were available as a competitive barrier?  Could the modern tech. revolution have happened without this competitive moat?

Very few of us would give up the incredible technological advances of the past 20 years.  But, these advances are helplessly intertwined with the development of these Closed Access policies.  And Closed Access policies invariably lead to incivility, inequality, and economic stress.  There are few things I feel more strongly about than that we should fight the tendency to impose Closed Access policies.  But, what if there is some tipping point with revolutionary innovations?  What if capital inflows to revolutionary innovations become uneconomical without a little bit of competitive protection?  If giving up Closed Access policies and their related social disorder meant that I would have to give up Yelp and Facebook and BlogSpot, is that a trade I would be willing to make?

It seems as if I am being extreme here.  Surely Closed Access policies aren't that important.  Surely most of the recent developments of frontier consumer technology are natural outgrowths of the current state of the art.

I have sort of been pushed to this idea through the back door.  Here is a chart from a speech (pdf) from Ben Bernanke comparing home prices during the boom to changes in each country's current account.  Now, Bernanke sees this as a defense of his monetary policies.  He sees a savings glut in the developing world as the causal factor bringing capital into the developed world, pushing down real interest rates, and thus pushing up property values.  But, that causation leaves many mysteries, including the first obvious question, "Why did that capital flow to Australia but not Germany?"  This is a similar question to the question we should ask about the US housing bubble.  "Why did predatory lending lead to home prices doubling in coastal California but not in Texas (where many of the new homes were being built, after all)?"

The answer to both of these questions is that supply is local.  Closed Access policies are local.  The causal factor here is Closed Access policies.  Firms located in these highly innovative economic centers that had competitive moats around them because of real estate limitations, could capture economic rents.  I am developing the idea that the capture of these rents is pulling down real incomes for the rest of the US.  When we click an ad on google, a micro-penny gets transferred, eventually, to the bank account of a landlord in San Francisco.  But, these firms are also capturing rents from the rest of the world.  Most of those excess profits are captured through foreign subsidiaries.  Bits don't have to be routed through the shipping terminals in LA, so they don't get recorded as international trade flows.  But, they do add to the value of those international subsidiaries.  This is why we have the mysterious situation where US assets held abroad consistently earn higher profits than foreign holdings of US assets, even though, year after year, foreigners invest billions more dollars in the US than we invest abroad.  This is because our corporations earn very high profits on their foreign activities which they continue to reinvest.  Foreign savers must continually send new capital to the US just to keep from falling behind.  And, to get the dollars to do that, they must sell us something.  The US trade deficit, at least in part, is a measure of the economic rents we earn from the rest of the world from our Closed Access housing policies.  The trade deficit isn't unsustainable.  We have already paid for those goods with the foreign profits of our corporations.

So, the causation starts at Closed Access policies.  This leads to both inflated home values and to excess foreign profits which lead to capital inflows as foreign nations that don't have access to these economic rents seek capital income to keep pace.

One of the tests of causation here would be to measure rents.  If the cause of these flows was a savings glut in the developing world, then that inflow of capital would trigger more homebuilding, so we would see falling rents and rising prices in countries with growing current accounts (inflows of capital).  If the cause was Closed Access policies, then we would see rising rents in those countries.  The rising rents would be the cause of both the rising home prices and the capital inflows, which are both simply the transfer of economic rents to the owners of Closed Access assets.  Much detailed work needs to be done here, but as I have shown in the US, the positive
 relationship of rents to home prices is very strong.  And, that seems to be the case in the international context, too.


  1. You wouldn't have to "give up the incredible technological advances." You'd simply have to use the next best thing. After all, technology often doesn't differ that much between competitors (uber vs lyft for example) but there is a network effect so you simply use what everyone else uses.

    1. Well, I think the effect might be the opposite. I wonder if now we have the next best version of more disruptive technologies, and if there weren't these competitive moats, we instead might have had more competitors fighting over improvements in less disruptive technologies....maybe. Speculative.

  2. This is (sort of) unrelated, but given the apparent social upheaval taking place in the US, i'd be curious to get your thoughts sometime on why the USA is so extreme vs other countries in mean vs media wealth per capita. Perhaps it's a good topic for a post. Seems timely - trade is taking all the heat but I doubt it's that simple...


    1. Hm. Some of the numbers in that table are surprising to me. Like the low number of millionaires in Hong Kong. For that matter, the US isn't so much high in millionaires as high in low wealth adults. According to that table. Is there a version from before the crisis to compare it to? That might be interesting.

    2. Page 4 has another version (but admittedly i havent spent any time with the source data).


    3. It would be interesting to compare it to numbers before 2007, because I wonder how much of the large number of low net worth adults are a result of the housing bust, which didn't happen in most countries.

  3. Why are high rents in San Francisco a moat around Yelp? Why can't entrepreneurs start Yelp competitors in Phoenix? Network effects? Are they really that large?

    One view might be as follows. Amazon, Facebook, Google, Yelp, DE Shaw, Two Sigma, SAC, AQR, etc were started relatively long ago. Perhaps costs of living in Closed Access areas were high but not insane then. Now these costs have reached insane levels, partly due to the enabling effects of higher incomes from these new firms. A common misconception about entrepreneurship is that one has an idea and starts by getting funding for it. Instead, usually, one slaves away for years with little to no income in order to build and prove out the idea and then raises funding as needed. With insanely high living costs and the low incomes associated with the early stages of entrepreneurship it becomes exceedingly difficult to access the network effects of the Closed Access cities. On top of this mechanism, operating and labor costs are higher in the Closed Access cities. But migrating out of them is also difficult because building a network of sources of labor and capital is difficult for founders. You can't just email a VC, you need to be introduced to them. This process can easily take months even for people with already extensive networks. Thus we see a slowdown in startup capital productivity. In the long term we should see "outsourcing" of startups to the Open Access cities, probably with much wailing and gnashing of teeth like in this Yelp story.

    But this moat around Yelp is at best temporary and the costs imposed on Yelp are high. Not only is there a high direct impact on costs but they are also draining their own talent pool. If the moat is temporary then it only postpones the inevitable by a few years. High initial startup valuations often rely on assumptions about cashflows that will happen at least a decade from now. This means the value of this moat is likely low. I would guess that in this picture Closed Access is likely a cost, not a benefit, to startups.

    VCs do, in fact, fund efforts to open new firms and offices of existing firms in low cost cities. New firms experiment with being located in lower cost suburbs that have easy access to city centers. Many firms also rely on telecommuting.

    It is entirely possible that firms like Yelp stay in San Francisco simply because the founders want to stay in San Francisco. See Noam Wasserman's "The Founder's Dilemmas", specifically the stuff on wealth-oriented versus control-oriented founders. I have no idea if Yelp really is this way or not. It also has always seemed to me that firms that like being in Closed Access cities tend to employ very young workers.

    1. I agree with you that collapse is inevitable. I think the rest of your comment is a great description of the nuts and bolts of the competitive moat and why Closed Access policies might have the ironic effect of geographically accumulating a certain type of firm.
      I think you need to take a step back and think about your interpretation. Your statement that "I would guess that in this picture Closed Access is likely a cost, not a benefit, to startups." is a confirmation of the value of the moat, not a contradiction of it. You are correct that it is a large cost to potential startups with long gestation periods. That is WHY it is so valuable to established firms, and why it is especially valuable to frontier tech firms.

    2. I'm sorry, I should have said it differently.

      I would guess that in this picture Closed Access is likely a cost, not a benefit, to typical established startups. This is because I think the value of this moat is low to typical established startups, likely lower than its cost. It is still possible to start a competitor to Yelp in Phoenix but it may take longer and take more effort. To justify a high valuation a firm must credibly be able to generate very high returns for a long time. My guess is that this time is substantially longer than the time it would take to breach this moat if the firm were to become vulnerable. Thus the existence of this particular moat has little impact on the valuation.

      It may depend on the point of the life cycle that the firm is in. Most of the famous tech startups are still investing heavily and not passing much cash to shareholders. For them the value of this moat may be low because their big dividends are still far in the future. You could say that they are not "established" but then "established" means IBM, and not even Google or Apple. Actually even IBM is investing heavily these days. Perhaps Yelp is different, I don't know.

      Even for "Value" firms as opposed to "Growth" firms this particular moat may not be that valuable because they have to hold their value for a long time. Regulatory capture, on the other hand...

      Perhaps the cost to early startups is larger than I think. It does seem like startup founders would be one of the most elastic groups when it comes to substituting untested stuff (offices in Phoenix) for costly stuff (offices in SF).

      Just speculating, obviously.