This is largely a school project I had a few weeks ago, and sadly it looks more relevant than ever. Let’s get it over with as soon as possible because this is some bitter stuff.
My entire discussion will be very basic, mostly to put things into perspective and make the model transparent, and also because I am not extremely well acquainted with the intricacies of the Congressional procedures and steps from here to debt ceiling increase, which is inevitably where we are headed. We begin with the crude GDP model:
GDP = G + I + C + X – M
G = Government spending, I = Fixed Capital Formation, C = Consumption, X = Exports, and M = Imports.
From here, lets plug in come numbers – the Bush Tax Cuts are around $300 billion, or 2% of US GDP. The spending cuts under the sequestration criteria are $150 billion annually, or 1% of GDP. Assumption: all the tax increases will be born by consumers (corporate tax rates are ridiculously low effectively in the US) and they will have a multiplier of 0.7 as consumers draw down savings to make up for this. Further assume that the spending cuts needs to come in discretionary income, and that the least sticky of these are mostly payouts from the state to individuals in different forms that the US government is not required to do by law (the cuts need to be made immediately, and I doubt Obama either can or wants to make cuts in staff) as well as deductions befalling non-lobby protected entities, and that the final multiplier for this is roughly 0.5.
We thus get 2 * 0.7 + 1 * 0.5 = 1.9. This is my base case of the percent GDP draw down immediately if the cliff edge is crossed. I would allow roughly 0.7% uncertainty to this given how little I know about the matters, and mostly how much these multipliers can change depending on the different resolution scenarios. The worst part is that the spending cuts need to be done, and they are thus not highly likely to be skirted, so the question then is how much of the tax increases will be taken on later.
From earlier mentioning, essentially that all of this 1.9% goes into consumer spending. Assume that imports are untouched as the substitution effect of American-for-foreign produced goods makes up for the lowered consumption. Thus, consumption if American-made goods would fall roughly 2.7% as the consumer is nearly 70% of the economy! (1.9 / 0.7 ~ 2.7).
Assuming that American goods and services are bought from companies listed in the Wilshire 5000 Index with a market cap of roughly $16.75 trillion, and that these companies are representing the vast majority of American corporate debt of $12 trillion, a scary picture emerges. First, the Wilshire 5000 is essentially financially levered 3.53 times to US GDP growth, although it is difficult to discern operational leverage. We assume that the S&P 500 is somewhat similar for the purpose of calculation, although I personally believe that the S&P 500 is relatively better at acquiring loans than it is at acquiring revenue. Say 3.5 times to make the number of significant figures equal, and then apply the full fiscal cliff. 2.7% * 3.5 = 9.45% Fundamentals for stocks in the US would deteriorate by 9.5% overnight! I’d say the S&P 500 is thus in for roughly a 12% earnings hit at least under the fiscal cliff. To estimate what this means in terms of free cash flow, we really need to look at operational leverage as well as financial, although this is more difficult. On top of this, I believe that the S&P 500 is richly priced at 14.4 P/E as it is right now. Running the data, I get a harmonic mean (mathematically correct way of averaging fractions) of 13.5! Even if I take the geometric average, which is slightly kinder, I get 14.4!
Yes, I know. The world is financialized, and we have leverage right now, so we can allow ourselves to be slack and use the normal, arithmetic average, right? Uhm… leverage only applies if you get growth, otherwise it works to your detriment! We are risking growth, so it should kick in reverse right now! I won’t model this, but just based on the harmonic mean using 12% lower earnings, we get almost exactly 1200! If we expect that the hit will bring the S&P down to one standard deviation below mean in harmonic terms… we get a P/E of 9.88! Using the earnings downgrade of 12%, we then get the S&P at 880!
The fiscal cliff is high for stocks!
For currencies, things look slightly better, though looking at the world through that prism, it is entirely reasonable that a hit like this would be comparable in structure to the credit crunch of 2009 initially, where GDP fell by around 4% from an initial growth of around 2%. Comparing the total 1.9% draw down from the growth we have at the moment, at a run rate around 2% for the last year, this would imply zero growth, and one-third to half of the severity of Lehman. The trough-to-peak increase was roughly 22% in the Dollar Index, and would imply 7%-11% increase currently, indicating a dollar index of 85.8 to 89.0 and by conversion a Euro at 1.2000 to 1.1712, although I do believe that the testing of the 1.20 level is much more likely. Likewise, the yen would fall to 88.3 to 91.6 versus the Dollar.
We will leave treasuries for now, they are largely powered by the incremental buying by the Fed.
Quantitative Easing! What does the Federal Reserve need to do here to save the day?
We get over to the adjusted Equation of Exchange: Inflation = ([Money Supply growth factor] * [Velocity of Money growth factor] / [GDP growth factor]) – 1
We now need some more data for this. Using a “worst-case scenario” and expecting that the Federal Reserve will not let inflation fall below 1% in a bid to keep deflation from occurring, here are some estimates. We use GDP growth at 1.4%, (1.014 in our formula) and look to the St. Louis Fed for some historical data on the fall of the velocity of money compared to a year ago. It seems the most likely indicator to watch for in a modelling perspective. Strangely, it also seems to follow a very specific pattern:
Moving downwards between 7.5% and 10% every shock!
We can thus expect velocity of money to slow down by at least 10% should the fiscal cliff be occurring, and likely hit the 12.5% slowdown state. Assume 10%, and then a further fall after that dragging the next measurement down to 12.5% slowdown under the scenario of the debt ceiling (more than I will discuss here). We plug 0.9 into the equation.
The only thing left now is the money supply, which the Federal Reserve controls. Calculating backwards, we get 1.01 * 1.014 = 0.9 * M2. Thus the Fed needs to quantitatively ease by increasing the money supply by 13.8%… of $10.25 trillion! Or total $1.41 trillion. Just to safeguard against deflation. Assume that the QE3 policy runs, and that we then use unsterilized purchases to suppress the long end of the curve under Operation Twist. This yields $40- + $45 billion per month. On an annual basis, that is $1.02 trillion – we would need another $400+ billion in QE besides what is already here just to avoid deflation!
Discount the roughly $500 billion (5% of money stock) already in QE3, and we get a further dilution by 9% of the Dollar. That brings the Dollar Index right back down here, if not even further! Of course, we are using these numbers “naked” now, and likely we will see the rise in the Dollar Index still – after all, that was what happened the last two times, and eventually when the disasters died down the only thing remaining was the dilution.
We are much more likely to see the effect in bond yields instead. Of course, they are very dependent on the methods the Fed uses, but just look at an example of the 10-year yields. The last times we had crises (Lehman, debt ceiling) 10-year yields fell 43%. Applying that to today’s 1.606% yield, we would arrive at 0.908% yield! Both these yield falls were eerily similar, that chart also showing the 38% fall after the start of the European debt debacle, albeit it took a significantly longer time. If there is any time the US is standing perfectly positioned to turn Japanese, this is it! Likely, it will get even worse now in “real” terms, since they will need to buy outside the 10-year Treasury market for effect, but this is looking less likely to materialize still by the minute. Remember, Qe1 was $800 billion, QE2 $600 billion, and QE3 about $480 billion per year. We seem to “need” $900 billion now!
That is likely the only way to keep stocks falling to the lowest rung of my analysis: debase the currency by 9%, making the S&P harmonic one-SD-below-mean case around 967, and the mean itself at 1320. Maybe the Mayans were onto something after all: the last day Congress can act in the session of the year happens to be December 21st, unless there is some unconventional recall from Christmas and New Year’s recess. If we go over, then lets discuss the possibility of the debt ceiling – it will be fun for the whole family (and world)!