Leverage and sticky prices — am I wrong?
RSJ, whose excellent blog is windyanabasis, takes issue with my claim that financial leverage is a source of sticky prices. Not only that, but he’s performed an impressive experiment to test and disprove the hypothesis. Check it out.
I am not persuaded that I am wrong, for reasons that I describe in a lengthy comment. But I am hardly an impartial reviewer. What do you think?
Update: RSJ has updated his experiment in response to some of my comments.
Update History:
- 21-May-2011, 5:15 a.m. EDT: Added bold update with link to RSJ’s follow-on post.
I don’t think you are wrong, but sadly, I don’t have time for a detailed rebuttal.
May 19th, 2011 at 12:16 pm PDT
link
Advantage Waldman at this stage.
May 19th, 2011 at 12:24 pm PDT
link
Quick observations on an interesting exercise:
First, capital structure matters. The two capital structure alternatives provide very different outcomes and opportunities if the firm doesn’t “meet its overall cost of capital”. Equity holders and debt holders (where they exist) partition this overall cost of capital into two distinct pieces. The leveraged firm runs the risk of getting “stopped out” from further equity participation (e.g. bankruptcy) when a shortfall in total absolute returns exhausts any return that remains available as a residual to the equity shareholders. Conversely, the all-equity firm has no such contractual “stop out” at the same level of absolute return shortfall. The rigid point of discontinuity that exists in the leveraged firm (comparable to a CDO “attachment point”) does not exist in the all equity firm. The equity holders continue to suffer from further downside risk at that point. But that may be a good thing for the survival of the firm longer term. There may well be consequences for not meeting the required overall return on capital in either case – at any stage of shortfall in returns – but the all equity firm offers much more flexibility in how the firm and its shareholders can respond along the way – including reducing prices. There is simply more downside risk protection against the rigid outcome of bankruptcy for shareholders in an all equity firm, notwithstanding the ever present risk of negative shareholder reaction to suboptimal management performance under any capital structure.
Second, the measurement of standard deviation in price changes seems suspect. SRW captures this point I think in his comment area where he talks about asymmetry. The potential pricing behaviour that one wants to investigate here is that of downward drift – not variance.
May 19th, 2011 at 2:34 pm PDT
link
SRW,
About dispersal within the industry, keep in mind that there are many more 5 digit NAICS codes than there are listed firms — each grouping is going to have only a couple of firms in it. Just think of “Snack Food Manufacturing”. How many of the top 1000 firms are in that group? We live in an oligopolistic economy with economies of scale.
The other snack food manufacturers — there *are* small time and specialty manufactures — will not have access to the credit markets. They will be funded by equity. At best, they might qualify for a loan, but the terms will not be favorable — outside the real-estate space and inventory financing, there isn’t a lot of long term bank lending to firms, particularly firms outside the top 1000.
By looking at data for firms that do have access to the bond market — (and really only the top few hundred have access) — you are going to skew the data towards market leaders, and you will see only a couple of firms in each industry. I’m not sure what value a statistical analysis would have on a sample size of one or two.
But this is the universe in which your theory would apply.
***
About the freedom that firms have — which constraint is actually binding?
Firms shut down plants and engage in mass layoffs all the time in response to a downturn in profitability. Even firms with zero debt. Next up, I’ll look at the correlation between mass layoffs and leverage.
***
About the game theory argument of whether a profit maximizing firm would lower its price, I don’t really like this framework of analysis because the firm is maximizing earnings — what is delivered to the owners of the firm — not profits in the economic sense. Managers do not get to keep any profits for themselves — all earnings are passed onto capital. Managers are hired to maximize the returns to capital, which are not zero in a competitive market.
If the equity investors as a whole expected earnings of $X — so that $X of earnings at the current price signals the indifference level of holding the stock. Now the firm delivers $X/2 — and it is expected that $X/2 is the new “permanent” earning.
The investors will sell the stock until its price falls to $P/2, and at that he point, they are back to their indifference level.
At least over the short term, the price of capital adjusts, rather than the required return — and that is true for an idiosyncratic shock as well as a systemic shock.
In the course of this adjustment, the owners of the firm are signaling to it that its return on investment is not allowed to decline. The firm (rationally) responds by reducing investment, and this results in less output and employment.
No need to invoke price stickiness. It doesn’t matter if the firm lowers its price and sells all of its inventory or not. What matters is that the firm will re-invest less.
***
The “rabbit hole” that I was referring to is the concept that there is some exogenous supply of real output available and all that is needed is the right market clearing price.
But if supply is itself a function of demand and vice versa, then you would look at the problem a bit differently.
In terms of measuring changes to aggregate output and employment, you don’t particularly care about whether consumer goods prices are sticky or not — you care about the return on investment demanded. Because capital goods prices are *not* sticky, reductions in expected profitability lead to a fall in capital goods prices (e.g. the enterprise value of firms) rather than a fall in capital rental rates. Capital rental rates appear sticky — and actually increase when earnings expectations decrease — and this would be true whether the financing is via equity or debt.
Which is not to say that there is no difference between failure to pay debt and failure to pay dividends. Only that this distinction is not a binding constraint and doesn’t show up in the aggregate data.
Good discussion!
I will post more data along the lines you pointed out a bit later on.
May 19th, 2011 at 5:35 pm PDT
link
Agree with JKH on both points, although I believe that the first one is the really the important one.
Debt financed firm has a time-horizon under which it cannot go, it will go bankrupt. It must meet its interest payments, full stop.
There’s no such restriction on an equity financed one.
Take an example of a tech startup – it can operate for years on no capital except for the equity invested by its employees/shareholders (you can work for nothing, even using your own home as an office, with your own computer & infrastructure). The same company, paying the employee wages and financed by debt would be bust very quickly.
Ergo there’s a capital structure difference.
On the prices analysis – as a matter of fact, I’d suggest that the stickiness may be in the margin (in absolute numbers, not percentage) rather than prices per se.
After all, it makes no difference to you if you sell a good at 100 and get to keep 20 out of it, or at 80 and get to keep 20 – your operating profit is the same.
Yes, of course, if you can hold the prices and get a higher margin you will do it. But then it makes sense for your compatitor to lower prices (and maintain margin).
We get to Bertrand competition scenario – whoever as lower production prices wins, if the production prices (for undifferentiated product) are the same, demand = capacity).
May 20th, 2011 at 5:29 am PDT
link
Another way of saying it:
For a given adverse shock, the shareholder in the all equity firm (compared to the shareholder in a leveraged firm) faces a greater expected loss, but a lower probability of bankruptcy. The lower probability of bankruptcy conveys greater inherent flexibility in dealing with issues of pricing (as well as other strategies) as it relates to longer term survival of the firm in its existing corporate form. That certainly doesn’t preclude the existing shareholders from losing on their investment, at least for a period of time – in the event of additional equity capital being raised at a lower market price, for example. Equity is risk capital, after all.
May 20th, 2011 at 8:10 am PDT
link
Let me preface this with a profession of my own distance from economics as a discipline, as a businessman I’m simply trying to find frameworks that maps to my experience. So the tone of what follows is blunt not because I take issue with your or RSJ’s judgements, vastly more expert than my own, but simply for the efficiency of getting to the point.
RSJ in abstract may or may not be right that debt and equity have equal affects, but a useful comparison should not be limited to sources of finance, between equity and debt, but extended to that between finance as a unit and cash flow. His argument uses the corporation as the minimum unit of concern. In that world he may be right and I have scant expertise to comment, your comment suggests he has some homework to do. But in the world of small privately held business, finance whether debt or equity, in my experience, is a cause of stickiness in price. Small business may not be as easy to study as publicly traded companies, but it remains a large part of the economy. It seems to me analysis like RSJ’s are a big part of why that part of the economy is at present choking: regulating an economy for the structure of its larger and more predictable entities could and I believe does have strong unintended effects in the unrecognized and understudied balance of he system.
His data analysis starts with the top 1200: out of how many companies? These top companies are the most likely to be managing with finance rather than cash flow, but data simply don’t exist for large parts of the small independent activity in the economy. The assumption that what is small behaves like what is large is like imagining our cells have consciousness because we do. He’s simply ignoring half the economy oblivious to the exclusion while you are picking nits within the parameters of his assumption.
If you haven’t read Sudhir Venkatesh’s “Off the Books” I encourage you to do so because in my experience most small business exist in an unexamined region between his black markets and RSJ’s financial ones. In the small business economy financial aphasiacs like me are the pricing function and “management” is compressed into both “ownership” and “labor”. Data points in business decisions are what information I can manage to organize into useful bits and my sense of what does and does not work. As a small business owner I find the world Venkatesh describes, or that of Gerd Gigerenzer in “Rationality for Mortals” much more realistic than most economic writing (this and several other web sites excluded), and the weight of scheduled commitments is the single largest weight a small business owner carries. This makes cash flow obligations, whether debt service or equity commitments a fundamental piece of the pricing equation. For what it’s worth.
May 20th, 2011 at 8:56 pm PDT
link
I updated the analysis to include:
*consolidating leverage by industry, so that we are comparing industry (5 digit) leverage to industry (5 digit) BLS producer price series
*I compared them to only standard deviations in downward movements in price, not all movements in price
*Weighted everything by size of industry
These were all good feedbacks. I appreciate other ideas for massaging the data to get more information out of it.
I still think there is no evidence of a relationship between leverage and downward price stickiness. There may be a relationship, but if so, you will need to regress along other parameters as well.
And I’m still puzzled as to why people are focusing on price stickiness as an explanation for output gaps. There was no price stickiness in Keynes’ original formulation, and you do not need price stickiness to argue that firms adjust quantities produced in response to changes in goods prices. There is already enough “real” stickiness in the existence of fixed non-circulating capital stocks that must be pre-funded that we do not need to attribute stickiness to prices.
I think the fact that we do is a measure of the poverty of our analysis, not a measure of any real world issue with prices.
With respect to the rights and obligations of equity holders versus debt holders — yes, but why is this relevant? Bankruptcy is not a binding constraint, firms change their production and investment plans in an effort to maintain their operating margins long before debt payments become an issue. I *think* at the macro level, these shifts in investment are what drives shifts in output and employment. I don’t think it matters very much whether a firm decides to lower prices of existing inventory or not if it is canceling the investment project or product line.
At most, price stickiness of existing inventories is a second order effect when compared to changes in production decisions for producing future inventories.
May 20th, 2011 at 11:41 pm PDT
link
[…] Firm Leverage and Price Stickiness, Part 2 Posted on May 21, 2011 by rsj This is a post in response to some of the issues raised by SRW. […]
May 20th, 2011 at 11:48 pm PDT
link