Archive
Half-Mast Isn’t Half Bad
As I watch the stock market, implausibly, rise to levels no one expected so soon after the crash, I am also sickened by the cheerleading from those whose fortunes – not to mention egos – are wrapped up in the level at which the Dow trades. Stock market fetishism always fascinates me as much as it repels me. Although my experience as a trader (and a short-term options trader, at that) would seem to suggest otherwise, my makeup is as a long-term investor. I want to buy value, and the mathematics of investing for me is that I want (a) high intrinsic value at (b) a low price. While people who are buy-and-hold investors of a certain age clearly benefit from higher prices, young investors clearly benefit from lower prices since they’re going to be net buyers for a long time. And the price at which you acquire intrinsic value matters. So does the price at which you sell, but not until you sell.
So to me, there’s nothing great about a price that’s high relative to intrinsic value, unless I am preparing to sell. In a broader sense, the idea that we should cheer for higher prices (as opposed to higher intrinsic values) is not only unseemly, but destructive and I’ll explain why in a minute. I will note that the fascination with watching prices ticking every second goes back a long ways: you can read in Reminiscences of a Stock Operator about the bucket shops of the early 1900s where speculators would watch and trade the stock market tape. The general increase in investor twitchiness and short-termism that has accompanied the growth of financial news TV, online investing, and the development of ETFs to trade broad market exposures intraday certainly adds numbers to the cheerleading crowd. But it isn’t new. Depressing, but not new.
The most fascinating example of this belief in the (non-intuitive, to me) connection between the value of the stock market and the value of ‘Merica was presented to me in the aftermath of 9/11. When we first trudged back to lower Manhattan, there were people handing out these cards:
Fight Osama! Buy Cisco! I never did see the connection, and it seemed to me at the time either delusional (the terrorists win if my investment in Lucent goes down) or nakedly self-serving. Certainly the way I feel about my country has nothing to do with where I’m able to buy or sell eBay today. And actually, they really weren’t coming for our 401(k)s, they were coming for our lives. But I digress.
The point I actually want to make is that when the Fed works to stabilize market prices, they’re having a negative effect by destabilizing economic variables. An analogy from manufacturing might be an entry point to this explanation: a truism in manufacturing is that you can stabilize inventory, or you can stabilize production, but you can’t stabilize both (unless your customers are accommodating and provide very smooth demand). If you want to stabilize inventory levels, then you need to produce more when business is high and less when business is low, so you’re on the production roller-coaster. If you want to keep production level, then inventory will be low when business is high and high when business is low, so you’re on the inventory roller-coaster. Only if business itself is stable – which is rare – can you do both.
A similar thing happens in capital markets. You can stabilize the cost of capital, but then you destabilize growth rates. Or you can stabilize growth rates, and the cost of capital (stock and bond prices) will fluctuate. This is true unless you can do away with the business cycle. If you choose, as the Fed has in recent years, to try and stabilize market prices at very high levels (stabilizing the cost of capital at very low levels), then when underlying activity is strong you’ll get a ton of speculative investment in capacity, new ventures that depend on the availability of cheap capital, and strong growth. And then when economic activity heads lower, you’ll find that lots of businesses go bust and the recession is deeper. In fact, it’s not just the speculative businesses that go bust, but the overbuilding in the expansion can cause even prudent enterprises to have more difficulties in the downturn.
The Fed’s historical response to this has been to let the speculative activity happen when the cost of capital is held too low, but not let companies go bust when economic activity wanes…so they lower the cost of capital even further.
I’m obviously not the first person to point out that the Fed’s constant intervention has deleterious effects and tends to increase the amplitude of boom and bust. And, for what it’s worth, I’m not blaming the current recession on the Fed. Clearly, the proximate cause of this recession was COVID-19 and the global economic shutdown. What made it worse was that the Fed, by holding down the cost of capital, had previously precipitated the development and preserved the success of many more speculative enterprises. And the fetishism about stock prices, and about how important it is to have lots of money “working for you” in the stock market, is also one of the reason that people don’t save enough.
Of course, right now is probably not the time for the Fed and Congress to pull back and let huge numbers of people and companies go bankrupt. There’s a case to be made for the sort of government response we are having in this episode, in which personal income is being replaced by money creation while workers are ordered to stay home. There will be a piper to pay for that policy – a loss of price stability which is a consequence of trying to preserve output stability, but a consequence that it’s arguably acceptable to pay. Afterwards, though, I hope that central banks can start to let natural rhythms replace the autocratic ones. I am not hugely optimistic on that score, but one place to start is this: stop lowering the bar for central banks to intervene in markets. Stop targeting equity prices and interest rates. It’s okay to let the Dow trade at half-mast, and the bull will come back without the Fed’s help.
In fact, if we don’t keep trying to artificially increase the length of the mast, the Dow might never need to trade down to half-mast in the first place. Certainly, intrinsic values don’t retrace 50% in a recession!
Summary of My Post-CPI Tweets (April 2020)
Below is a summary of my post-CPI tweets. You can (and should!) follow me @inflation_guy. Or, sign up for email updates to my occasional articles here. Investors with interests in this area be sure to stop by Enduring Investments (updated site coming soon).
- CPI Day! I have to be honest – with the markets closed and this number likely to not have a lot of meaning, I almost skipped doing this this morning. But, thanks to M2, lots more people are suddenly listening so…obviously CPI is starting to be important. So let’s try.
- The consensus for today, to the extent “consensus” means anything, is for -0.3% headline and +0.1% on core. But these are even more guesses than usual.
- The BLS stopped taking prices a couple of weeks ago. That will have less effect than if that had happened a few years ago, b/c they ‘survey’ some prices using database downloads from retailers e.g. apparel.
- But it still means that we don’t know what they’ll do about missing prices. Normally the BLS imputes an estimated figure for an item based on similar items…but if whole groups of items or categories are missing, less clear. Do they assume zero? Prior trend?
- I actually think that this number won’t have too many of those problems but there will be some and next month will be very odd – and some chance they don’t publish at all because they can’t get statistically significant data.
- In the meantime…remember we are coming off of recent strong data. Core was 2.37% y/y last month, and in general has headed higher. Before this, I was expecting 2.5% core by summer. Now that will take longer! (You can imply the word “maybe” before every statement this month.)
- Lodging Away from Home is one place we’ll surely see an effect this month, and airfares, but beyond that who knows. And we are dropping off a weak +0.16% from last March so the core y/y figure might even stay steady. Or it could drop 0.3%. Who knows.
- What we DO know is headline inflation is going to fall and in a month or 2 will show negative changes, which will prompt “DEFLATION!” screams. But headline just follows gasoline. That’s important – it’s also the reason people think infl is related to growth. Only headline.
- I doubt we’ll get very close to core deflation in this cycle. See my recent article Last Time Was Different for why I don’t think we’ll see similar effects. But mkts are priced for long-term disinflation and deflation.
- Oh and of course yesterday’s M2 chart. Probably discuss that more later today. Anyway, I’d say good luck but with markets closed you can’t do anything anyway! So just “hang on” and we’ll try and figure this out over the next few months.
- I will be back in 5 minutes with thoughts on the figures and diving as deep as I can this month.
- Core -0.1% m/m, down to 2.1% y/y. That’s a bigger fall than expected, but with these error bars I wouldn’t be shocked. Normally missing by 0.2% on core is a big deal. More interesting is that they got headline right to within 0.1%! It ‘only’ fell -0.4% m/m in March.
- Here are the last 12 core CPI prints. This chart is gonna look kinda wacky for a while.
- Broadly, core goods were -0.2% y/y, a decline from flat. More amazing is core services, dropping to 2.8% y/y from 3.1%.
- Haha, that core services number is EVEN MORE AMAZING than you think. Because it didn’t happen from Owners Equivalent Rent (+0.26% m/m, 3.22% y/y) or Primary Rents (+0.30% m/m, 3.67% y/y). Both slower y/y but basically same m/m from Feb.
- So if rents didn’t decelerate, where do we get the big drop in core services? Lodging Away From Home was -6.79% m/m, dropping to -6.38% y/y from +0.78% last month. I should drop the second decimal.
- BTW, good time to remember that VOLUMES of transactions don’t enter into CPI monthly. This is just a survey of prices. So if no one bought any apparel, but we have a price, that’s what gets recorded. Lodging fell because prices actually were down hard, as you probably know.
- CPI for Used Cars and Trucks was +0.82% m/m. Some people were worried about autos but I’m not sure they should be. Big supply shock in cars because of parts supply chain. If I were a dealer I wouldn’t be marking down my existing inventory.
- Airfares -12.6% m/m. That’s worth about 0.1% on core all by itself. So we expected big declines in airfares and Lodging Away from Home (worth about 0.06%), and got them. Core ex- those two items still had some softness, but not horrendous.
- Core ex-housing declined from 1.70% y/y to 1.45% y/y. Again, a lot of that were those two items I just mentioned. But 1.45% core ex-housing is still higher than it was last July.
- Now, in medical care I’m not sure how to think about any of this. Medicinal Drugs were -0.04% m/m, after -0.43% last month, pushing y/y to 1.31% from 1.85%. But lots of drugs are really hard to get right now and of course we now know most of our APIs come from China.
- That may be a case of some shortages, because in the short term no one wants to be seen jacking up the price of drugs. Prescription drugs decelerated y/y; non-prescription accelerated.
- Physicians’ Services +0.34% m/m vs +0.21% prior month. Hospital Services +0.40% vs -0.12%. How in the heck do you measure this when most of those doctors and services are doing one thing? And a very crucial one indeed. What’s the price of a hip replacement right now?
- OK, biggest m/m changes down, other than fuel. Public Transportation -65% (annualized), car/truck rental -58%, Lodging Away from Home -57%, Infants/toddlers apparel -41%, womens/girls apparel -30%, footwear -29%.
- Which makes me realize I forgot to mention Apparel was -2% m/m. That’s another 5bps off the core inflation rate.
- There were still some increases on the month. Biggest ones other than food were Tobacco and Smoking Products (12.5% annualized), nonalcoholic beverages (+12%), and Used Cars and Trucks (+10%).
- FWIW, the early look to me is that MEDIAN CPI will still be around 0.22% or so. That’s what long-tail negatives do to core! So while y/y Core dropped sharply, y/y median will still be around 2.8%.
- So, coarse but…core -0.1% m/m. Add back 0.06% lodging, 0.10% airfares, 0.07% apparel and 0.07% for public transportation (cuffing it) and you get back to +0.2%. Which means that outside of those categories there wasn’t much disinflation pulse. Median will say same thing.
- That probably more means that prices haven’t really reacted yet that that there will be zero impact of COVID-19. But I don’t think we’ll see a big impact lower on prices. At least not lasting very long.
- Haven’t done many charts yet. But here’s one I haven’t run in a while. Distribution of y/y price changes by low-level item categories in the CPI. Look at that really long tail to the left. Take off just the last bar on the left and you get 2.37% core roughly.
- Here’s the weight of categories over 2% y/y change, over time. Just another way of saying that we haven’t seen any big effects yet. Unknown is just how much the trouble in collecting affects this.
- Pretty good summary and gives me more confidence in the data – they’re at least calling people! But interestingly, not so much doctors/hospitals. So asterisk by Medical Care.
- BLS has posted this, explaining how they’re collecting prices. https://bls.gov/bls/effects-of-covid-19-pandemic-on-bls-price-indexes.htm#CPI
- So let’s do the four-pieces charts and then wrap up. For those new to my monthly CPI tweets, these four pieces add up to CPI, each is 20%-33%, but each behaves differently from a modeler’s perspective.
- First piece: Food and Energy. This will go much lower. As I said up top, we will be in deflation of the headline number pretty soon. But, I think, only the headline number.
- Core goods. This declined a tiny bit, mostly apparel. I think the short-term effect here is indeterminate but might actually be higher as some goods made overseas get harder to get (ibuprofen??)
- Here’s where the rubber meets the road. Core Services less Rent of Shelter. Was in a good trend higher and about to be worrisome. Dropped a bit, but with an asterisk on medical care.
- Rent of Shelter – this looks alarming! And rents declining is the ONLY way you can get core deflation. But…Rent of Shelter includes lodging away from home. That’s the dip, is in that 1% of CPI. The 31% that is primary and OER, not so much.
- That last chart calls for one more on housing. Here is OER, the biggest single piece of CPI. It’s right on model. As yet, no sign of any big effect from COVID-19 either now, or in the forecast that’s driven by housing market data.
- End with 1 final chart. We started w/ M2 chart showing the biggest y/y rise in history. The counterpoint is “what if velocity falls.” But vel is already @ record low. To drop, you need lower int rates (from 0?), or huge long-lasting cash-hoarding.Hard to see.
- Thanks for tuning in. I’ll collate these in a single post in the next hour or so.
So what was most amazing about today’s data? I suppose it was that, outside of the things we knew would be disasters (airfares, hotels) the effects of the virus crisis were very small. And you know, that sort of makes sense. If I’m a producer of garden rakes (I honestly just pulled that out of the air), why would I change my prices? I’m not seeing traffic, but it isn’t because my prices are too high. From a seller’s perspective, it only makes sense to lower price if lower prices will induce more business. Lowering the price of rakes isn’t going to sell more rakes. It isn’t that people have no money to buy rakes – with the government fully replacing wages of laid off workers, and covering the wage costs for small businesses so they don’t need to lay anyone off, and sending everyone a fat check besides, there’s no shortage of people with money to spend. (I know we read a lot about the tragedy of the millions being laid off, but it’s not much of a tragedy yet since they’re being paid the same as before!)
[As an aside, businesses with high fixed overhead and low variable costs – hotels are a classic example; it costs very little for the second occupied guest room – might lower prices significantly since if they can cover their variable costs then anything above that goes to covering fixed overhead. That’s what airlines did initially too, but when they realized after that knee-jerk response that they couldn’t fill the planes even if they offered free flights, they started canceling enormous numbers of flights. I’ve actually seen some of the fares that I track rise in the last week or two as the number of flights out of NYC has dwindled to very few! But it’s harder to mothball a hotel than to mothball a plane.]
The NY Fed published a really insightful article today entitled “The Coronavirus Shock Looks More like a Natural Disaster than a Cyclical Downturn.” Although they focused on the path of unemployment claims, a similar analysis can take us to the inflation question. In a natural disaster, we don’t see deflation. If anything, we tend to see inflation as some goods get harder to acquire. The amount of money available doesn’t decline, assuming the government deploys an emergency response that includes covering non-insured losses, and the amount of goods available drops. In today’s circumstance, we have more money available – as the M2 chart shows – than we did before the crisis, and if anything we will have fewer things to buy when it’s all over as supply chains will remain disrupted for a long time and a lot of production will surely be re-onshored. But you don’t need the latter point to get disturbing inflation. All you need is for the money being created to get into circulation rather than reserves (which is what is happening, which is why M2 is soaring), and for precautionary money-hoarding to be a short-term phenomenon. I believe the money will be around long after the fear has died away, because for the Fed to drain a few trillion by selling massive quantities of bonds is much, much more difficult than to add a few trillion by buying bonds that the Treasury coincidentally needs to sell more of right now.
The quality of the CPI numbers will be sketchy for a while, but I am fairly impressed that this release wasn’t as messy as I was prepared for. The inflationary outcome may well be messy, though! With 14% money growth, and little reason to expect a lasting velocity decline, it’s hard to get an innocuous inflation outcome. But markets are still offering you inflation hedges at prices that imply you win even if inflation drops a fair amount from the current level. If you don’t have those hedges, you’re making a very big bet on deflation.
Happy Easter.
Last Time Was Different
They say that the four most dangerous words in investing/finance/economics are “This time it’s different.”
And so why worry, the thinking goes, about massive quantitative easing and profound fiscal stimulus? “After all, we did it during the Global Financial Crisis and it didn’t stoke inflation. Why would you think that it is different this time? You shouldn’t: it didn’t cause inflation last time, and it won’t this time. This time is not different.”
That line of thinking, at some level, is right. This time is not different. There is not, indeed, any reason to think we will not get the same effects from massive stimulus and monetary accommodation that we have gotten every other time similar things have happened in history. Well, almost every time. You see, it isn’t this time that is different. It is last time that was different.
In 2008-10, many observers thought that the Fed’s unlimited QE would surely stoke massive inflation. The explosion in the monetary base was taken by many (including many in the tinfoil hat brigade) as a reason that we would shortly become Zimbabwe. I wasn’t one of those, because there were some really unique circumstances about that crisis.[1]
The Global Financial Crisis (GFC hereafter) was, as the name suggests, a financial crisis. The crisis began, ended, and ran through the banks and shadow banking system which was overlevered and undercapitalized. The housing crisis, and the garden-variety recession it may have brought in normal times, was the precipitating factor…but the fall of Bear Stearns and Lehman, IndyMac, and WaMu, and the near-misses by AIG, Fannie Mae, Freddie Mac, Merrill, Goldman, Morgan Stanley, RBS (and I am missing many) were all tied to high leverage, low capital, and a fragile financial infrastructure. All of which has been exhaustively examined elsewhere and I won’t re-hash the events. But the reaction of Congress, the Administration, and especially the Federal Reserve were targeted largely to shoring up the banks and fixing the plumbing.
So the Federal Reserve took an unusual step early on and started paying Interest on Excess Reserves (IOER; it now is called simply Interest on Reserves or IOR in lots of places but I can’t break the IOER habit) as they undertook QE. That always seemed like an incredibly weird step to me if the purpose of QE was to get money into the economy: the Fed was paying banks to not lend, essentially. Notionally, what they were doing was shipping big boxes of money to banks and saying “we will pay you to not open these boxes.” Banks at the time were not only liquidity-constrained, they were capital-constrained, and so it made much more sense for them to take the riskless return from IOER rather than lending on the back of those reserves for modest incremental interest but a lot more risk. And so, M2 money supply never grew much faster than 10% y/y despite a massive increase in the Fed’s balance sheet. A 10% rate of money growth would have produced inflation, except for the precipitous fall in money velocity. As I’ve written a bunch of times (e.g. here, but if you just search for “velocity” or “real cash balances” on my blog you’ll get a wide sample), velocity is driven in the medium-term by interest rates, not by some ephemeral fear against which people hold precautionary money balances – which is why velocity plunged with interest rates during the GFC and remained low well after the GFC was over. The purpose of the QE in the Global Financial Crisis, that is, was banking-system focused rather than economy-focused. In effect, it forcibly de-levered the banks.
That was different. We hadn’t seen a general banking run in this country since the Great Depression, and while there weren’t generally lines of people waiting to take money out of their savings accounts, thanks to the promise of the FDIC, there were lines of companies looking to move deposits to safer banks or to hold Treasury Bills instead (Tbills traded to negative interest rates as a result). We had seen many recessions, some of them severe; we had seen market crashes and near-market crashes and failures of brokerage houses[2]; we even had the Savings and Loan crisis in the 1980s (and indeed, the post-mortem of that episode may have informed the Fed’s reaction to the GFC). But we never, at least since the Great Depression, had the world’s biggest banks teetering on total collapse.
I would argue then that last time was different. Of course every crisis is different in some way, and the massive GDP holiday being taken around the world right now is of course unprecedented in its rapidity if not its severity. It will likely be much more severe than the GFC but much shorter – kind of like a kick in the groin that makes you bend over but goes away in a few minutes.
But there is no banking crisis evident. Consequently the Fed’s massive balance sheet expansion, coupled with a relaxing of capital rules (e.g. see here, here and here), has immediately produced a huge spike in transactional money growth. M2 has grown at a 64% annualized rate over the last month, 25% annualized over the last 13 weeks, and 12.6% annualized over the last 52 weeks. As the chart shows, y/y money growth rates are already higher than they ever got during the GFC, larger than they got in the exceptional (but very short-term) liquidity provision after 9/11, and near the sorts of numbers we had in the early 1980s. And they’re just getting started.
Moreover, interest rates at the beginning of the GFC were higher (5y rates around 3%, depending when you look) and so there was plenty of room for rates, and hence money velocity, to decline. Right now we are already at all-time lows for M2 velocity and it is hard to imagine interest rates and velocity falling appreciably further (in the short-term there may be precautionary cash hoarding but this won’t last as long as the M2 will). And instead of incentivizing banks to cling to their reserves, the Fed is actively using moral suasion to push banks to make loans (e.g. see here and here), and the federal government is putting money directly in the hands of consumers and small businesses. Here’s the thing: the banking system is working as intended. That’s the part that’s not at all different this time. It’s what was different last time.
As I said, there are lots of things that are unique about this crisis. But the fundamental plumbing is working, and that’s why I think that the provision of extraordinary liquidity and massive fiscal spending (essentially, the back-door Modern Monetary Theory that we all laughed about when it was mooted in the last couple of years, because it was absurd) seems to be causing the sorts of effects, and likely will cause the sort of effect on medium-term inflation, that will not be different this time.
[1] I thought that the real test would be when interest rates normalized after the crisis…which they never did. You can read about that thesis in my book, “What’s Wrong with Money,” whose predictions are now mostly moot.
[2] I especially liked “The Go-Go Years” by John Brooks, about the hard end to the 1960s. There’s a wonderful recounting in that book about how Ross Perot stepped in to save a cascading failure among stock brokerage houses.
