“Brazil in Drag”: Hyman Minsky on Donald Trump

Via Nathan Cedric Tankus, here is a recent JPKE article by Kevin Capehart on a 1990 lecture by Minsky that uses Trump as a case study of asset market bubbles in the 1980s. The lecture is fascinating, and not just as an odd historical artifact.

Here is what Minsky says about Trump:

One of the puzzles of the 1980s was the rapid rise in the financial wealth of Donald Trump, author of The Art of the Deal… Trump’s fortune was made in real estate. Many large fortunes have been made in real estate, since real estate is highly leveraged. Two factors made Trump somewhat unique — one was the he developed a fortune in the period of high real interest rates, and the second was that the cash flows on most of Trump’s properties were negative.

Trump’s wealth surged because the market value of his properties — or at least the appraised value — was increasing faster than the interest rate. Trump obtained the funds to pay the interest on his outstanding loans by increasing the draw under what in effect was a home equity credit line. The efficiency with which Trump managed these properties was more or less irrelevant — hence Trump could acquire the Taj Mahal in Atlantic City without much concern about the impacts on the profits of the two casinos he already owned. Trump was golden — he had a magic touch — as long as property prices were increasing at a more rapid rate than the interest rate on the borrowed funds.

The puzzle is that the lenders failed to recognize that the arithmetic of his cash flows was virtually identical with that of the developing countries [discussed earlier in the lecture]; in effect Trump was Brazil in drag. In the short run Trump could make his interest payments with funds from new loans — but when the increase in property prices declined to a value below the interest rate, Trump would become short of the cash necessary to pay the interest on the outstanding loans.

The increase in U.S. real estate prices in the 1980s was regional, and concentrated in the Northeast and in coastal California. … Real estate prices dipped in the oil patch, climbed modestly in the rust belt, and surged in those areas that benefitted from the rapid increases in incomes in banking and financial services — sort of a derived demand from the financial success of Drexel Burnham. In effect, those individuals with high incomes in financial services — and with the prospect of sharp increase in incomes — set the pace for increases in real estate prices.

Trump’s cousins were alive and well and flourishing in Tokyo, Taipei and Seoul especially in the second half of the 1980s. The prices of equities and real estate were increasing because they were increasing…

In any market economy the price of real estate will tend to reflect both its rental return and the rate of return on the riskless bond. … The price of land rises and the price of land sometimes falls — the relevant question is whether the anticipated increase in the price of land is sufficiently higher than the interest rate on bonds to justify a riskier investment.


The key question is why so many varied bubbles developed in the last several decades. The most general answer is that sharp changes in inflation rates and interest rates led to extremely volatile movement in asset prices. And once these price movements begin, then on occasion momentum may develop and feed on itself — at least for a while.

So in Minsky’s version of The Art of the Deal, there are three things you need to get rich like Trump. First, be an investor in NYC and New Jersey real estate in a period when land prices are rising rapidly there relative to the rest of the country. Second, be highly leveraged. And third — and this is critical — convert your equity to cash as quickly as possible to protect yourself from the post-bubble fall in prices. Picking the right individual properties doesn’t matter so much, and managing the properties well doesn’t matter at all.

In this analysis, the repeated bankruptcies of Trump-controlled properties don’t undermine his claims of business success, nor are they just an incidental footnote to it; they are an integral part of how he got so rich. Because the flipside of extracting cash from his properties through “what was in effect a home equity credit line” is that there was less equity left for the entity that actually owned them.

The trick to making money in an asset bubble is to cash out before it pops. Doing this by selling at the peak is hard; you have to time it just right. It’s easier and much more reliable to cash out the capital gains as they accrue; that just requires some way of moving them to a different legal entity. The precedent for Trump, in this reading, would be the utility holding companies that played such a big part in the stock market boom of the 1920s and were such a big target for regulation in the 1930s. Another parallel would be today’s private equity funds. To the extent that the funds cash out via so-called “dividend recapitalization” (special dividends paid by the acquired company to the PE fund) rather than eventual resale, an acquired company that doesn’t end in bankruptcy is money left on the table. It’s interesting, in this context, to think about Romney and Trump as successive Republican nominees: They may embody different cultural stereotypes (prissy Mormon patriarch vs womanizing New York vulgarian) but fundamentally they are in the same business of financial value extraction.

6 thoughts on ““Brazil in Drag”: Hyman Minsky on Donald Trump

  1. You can say the same for any broker or wealth manager , most of whom have never beat the S&P 500 , but keep telling investors how brilliant they are

  2. ” To the extent that the funds cash out via so-called “dividend recapitalization” (special dividends paid by the acquired company to the PE fund) rather than eventual resale, an acquired company that doesn’t end in bankruptcy is money left on the table.”

    I don’t agree with this point at all. It shows a misunderstanding of the corporate lending and bond markets. It’s not possible to lever a company up to 100% of its estimated enterprise value, so a sale will at any given point net more than a dividend recap. That’s the basic math.

    Real estate lending is a bit different because it’s higher LTV with more emphasis on appraised value and less emphasis on cash flow.

    That analysis also ignores some meaningful other limitations. Bankruptcy courts reclaiming preference payments to related parties are a risk. So are lawsuits arguing that a business was insolvent immediately post-recap. It also ignores that PE firms and lenders are repeat players in these markets. Follow this idea too many times, and see how much debt is available for such a PE firm’s future transactions.

    1. I always feel skeptical when people say things are “impossible” based on this kind of a priori reasoning, or assume that various exploits are prevented by reputation costs. Besides, it’s not necessarily a bad deal for the lenders — the real gains come from reneging on the implicit and explicit (e.g. pension) contracts with other stakeholders. The bankruptcy court thing seems highly speculative — can you give me an example of a court actually reclaiming dividend payments? I find it hard to imagine.

      More fundamentally, dividend recaps do happen, and they are quite large. Debt-financed special dividends to private equity funds account for close to 10% of all dividend payments in the US in recent years. ($70 billion out of $900 billion in 2013.) There is no question that these payments make the acquired company financially weaker. You can say “oh no, no one would make these loans if they materially increased the chance of bankruptcy” but that’s just wishful thinking. Lenders are desperate for yield, of course they will.

      1. Thank you for your reply.

        The issue is not just one of reputation costs. Reputation costs are an additional argument beyond my primary one, which is the math of returns.

        The average equity contribution to LBO’s has been around 40% in recent years, with annual averages ranging between 33% and 51%. ( See the S&P LCD data on page 7 here – https://www.williamblair.com/~/media/Downloads/Emarketing/2015/IB/Lev_Fin_2015_05.pdf ) That has a couple important implications.

        First, and most importantly, it implies that a dividend generates substantially less proceeds than would be realized by selling the company. The average 2014 leverage multiple and equity contribution calculate to a selling multiple of 8.7x EBITDA (5.3x leverage multiple plus 3.4x equity implied by a 39% equity contribution). By comparison, the average post-dividend leverage was 4.9x. The difference in equity proceeds is even greater than it might appear at first glance because pre-dividend leverage is greater than 0. (This math is for a sale to another private equity firm. That’s sort of by definition likely to be a floor valuation in evaluating other options, namely sale to a strategic buyer or an IPO.)

        Second, the level of initial equity contribution also has implications for the magnitude of dividends required to generate an equity return. Pencil out the math of buying a company at around 8-9x EBITDA with around 5x EBITDA debt, then growing EBITDA at mid-single digits per year, paying down some debt from company cash flow, and periodically taking out dividends via dividend recaps at around 5x EBITDA. Then compare that to selling after 5 years, perhaps with a debt-financed dividend recap after 2 or 3 years.

        As for the idea that bankruptcy isn’t a bad outcome for lenders, the average ultimate recovery rates in bankruptcy from Moody’s database contradict that notion. (See http://efinance.org.cn/cn/FEben/Corporate%20Default%20and%20Recovery%20Rates,1920-2010.pdf ). The average recovery rates on loans from 1987-2010 are 80.3% (p. 7). As one would expect, these drop as one moves down the debt capital structure to just under 50% for senior unsecured bonds and under 30% for subordinated bonds. Those are hefty losses – especially when one remembers that these are instruments with on the order of 5% to 15% annual interest payments with no upside beyond principal and interest. (The upper end of that interest rate range isn’t really in play in the current interest rate environment, but it was for much of the time period cited.) And also when one remembers that these recoveries are after a Chapter 11 process that requires a fair amount of time and money on the part of lenders, so some sell out to distressed funds and only achieve the recovery rates – 66% for first lien bank loans – shown on p. 5 of this report.

        As for the risk of returning dividend payments, KB Toys and Powermate are two examples where lawsuits and settlements occurred. (See http://www.lexology.com/library/detail.aspx?g=ffc711dc-a0a7-4c7a-8a81-d014d6cb97d3#dividend. As an aside, the introduction is an interesting reminder of just how negative sentiment was in early 2009, as the predicted 50% default rate for portfolio companies obviously didn’t happen.)

        There are a couple of well-established principles for bankruptcy courts scrutinizing such payments. One is preference payments, where for insiders/related parties, the trustee can look back up to 1 year prior to filing. (See http://www.moranlaw.net/preferences.htm) The other is a “constructively fraudulent” transfer, where the bankruptcy code looks back up to 2 years prior to filing and some state laws provide for longer look back periods. Note that despite the use of the word “fraud”, intent to defraud creditors (“actual fraud”) isn’t required to reclaim payments. The focus is on (i) whether fair consideration was received by the company and (ii) whether the company was insolvent or left with unreasonably small capital after the transaction. (See http://www.law360.com/articles/553894/understanding-fraudulent-transfers-and-ensuing-litigation)

        So there are both practical/mathematical reasons – i.e., available financing – and legal reasons why pursuing the strategy that you propose is very difficult. Perhaps “impossible” is too strong a term, as there are always rare outlier examples in any market. I was reacting, however, to a categorical statement in the opposite direction, which doesn’t hold up based on the factors that I’ve cited.

  3. I assume this describes just the 80s era Trump, no? By most appearances, he’s about as legit as it gets by now: Landlording, Golf courses with good cash flow, Licensing his brand, etc.

    In fact, most of his money seems to have been made through normal business management. The Minsky lecture in 1990 was before Trump went bully-up. That game worked for him for a while, but he got caught over-leveraged. Contra Minsky, he didn’t get his equity out, was way underwater, and had to negotiate w/ his debtholders, who could have drove a stake in him if they wanted to.

Leave a Reply

Your email address will not be published. Required fields are marked *