My Inflation Outlook as of 2022-04-12 We; & BRIEFLY NOTED: For 2023-04-12 We
braddelong.substack.com
My Inflation Outlook as of 2022-04-12 We; & BRIEFLY NOTED: For 2023-04-12 We
The bond-market canary's plumage is still lustrous!; secular stagnation is likely to return; John Stuart Mill & Harriet Taylor; avoiding neighborhood blight via immigration; Tressie McMillan Cottom &
The 5-year 5-year-forward inflation breakeven—the inflation rate at which TIPS returns match nominal Treasury returns from 5 to 10 years from now—continues to hang at 2.2%/year, which roughly corresponds to a PCE chain-index inflation rate of 1.7%/year over those five years. The bond market’s marginal traders are thus anticipating that the Fed will be undershooting its 2.0%/year PCE-inflation target starting in 2028:
The five-year inflation breakeven is now at CPI 2.3%/year, corresponding to a marginal bond-market trader expected-inflation rate for the PCE chain index of 1.8%/year:
Inflation may well overwhelmingly be a monetary phenomenon. But central banks only choose to allow it to happen under one condition: when they sense the expectations of inflation are sufficiently firmly entrenched that a failure to accommodate those expectations with monetary expansion would produce deflation and depression more painful than the inflation. Thus behind any accommodating monetary expansion lurks entrenched expectations of continued inflation. And the canary in the coal mine with respect to expectations of inflation has always been the bond market. But the bond-market canary has not drop dead. The bond-market canary is not even sick. Its plumage is lustrous and full. And it is chirping away—the exact opposite of the scenario that would generate rational fears of stubborn sticky persistent inflation. The three sources of inflation are strong demand pressure, bottlenecks and supply shortages, and expectations. We have confidence that central banks will curb demand pressures, we have no expectational component, and bottlenecks and supply shortages appear under control.
And yet we have a very large number of very sharp commentators very worried about persistent inflation. I have gathered six representative examples below. Lars Feld believes central banks are greatly under estimating how much demand pressures they are currently are. Jim, O'Neil sees bottleneck pressure, emerging from Russia and demand pressure emerging from China. Mohamed El-Erian sees a good case in which services inflation remains sticky and constant and drives aggregate inflation to a stable 3 to 4%, and a bad case in which China and Russia drive inflation higher. Paul de Grauwe & Yuemei Ji seek to arm central banks with better tools and claim that it is time to begin massive quantitative tightening. Nouriel Roubini sees inflation emerging from the enormous spending demands that the world’s problems and voters’ preferences will impose on governments, coupled with those same voters’ reluctance to pay higher taxes, and thus anticipates the return of the fiscal theory of the price level. And Michael Spence sees a future world struck by bottleneck after bottleneck, adverse supply shock after supply shock:
Lars P. Feld (Mar 31, 2023): Why Inflation Persists: ‘Contrary to what many commentators would like to think, high inflation today is a macroeconomic phenomenon, resulting from too much money chasing too few goods and services. Worse, there are many reasons to worry that it will prove to be more persistent than many market participants have acknowledged…. Some… warn that monetary-policy tightening… risks overshooting. Such warnings are premature, if not counterproductive. Inflation tends to become persistent because of second-round effects. Not only have producer prices risen much more during this period, but they also have not yet been fully passed through the value chain…. Focusing on the relative strength of price components (like energy prices) as a causal factor overstates the microeconomic foundation of macroeconomics. Inflation still is what it always has been: a macroeconomic phenomenon that results from the disequilibrium of aggregate demand and aggregate supply… influenced by monetary and fiscal policy...
Jim O'Neill (Mar 9, 2023):The Inflation Picture Gets Murkier: ‘Despite signs that inflation in most major economies has peaked and is trending back down, recent data releases have renewed fears that central banks will have to tighten monetary policy still further. In the face of maddeningly mixed signals, much will depend on a few key factors…. Russia’s war in Ukraine remains a source of deep global uncertainty. At this point, no one has any idea when or how the fighting will end, let alone what economic consequences it will have, especially on commodity prices. The second source of uncertainty is China’s economic recovery…
Mohamed El-Erian (Feb 9, 2023): There Is More Inflation Complexity Ahead: ‘As US inflation gradually eases, the claim that today’s inflationary pressures are the result of a temporary supply shock has re-emerged. While this thesis may be comforting, it could also encourage dangerous complacency, making an already serious problem much harder to solve…. Three possibilities…. “Immaculate disinflation.”… Inflation becomes sticky…. at 3-4% over the second half of this year as goods prices stop declining and services inflation persists. This would force the Fed to choose between crushing the economy to get inflation down to its 2% target, adjusting the target rate to make it more consistent with changing supply conditions, or waiting to see whether the US can live with stable 3-4% inflation…. Lastly… “U inflation”: prices head back up late this year and into 2024, as a fully-recovered Chinese economy and the strong US labor market simultaneously drive persistent services inflation and higher goods prices. I would put the probability of this outcome at 25%…
Paul de Grauwe & Yuemei Ji: (Feb 7, 2023): Tame Inflation Without Subsidizing Banks: ‘Major central banks face an overabundance of reserves, owing to years of quantitative easing…. According to conventional wisdom, the only way central banks can push up rates in such an environment is by paying interest on the ocean of reserves held by credit institutions…. But there are other ways for a central bank to drive market interest rates higher without transferring its profits to commercial banks. For example, it could sell government bonds, a form of quantitative tightening that major central banks are already implementing…. Minimum reserve requirements… transforming the excess reserves held by commercial lenders into required reserves, on which no interest is paid…. Some object to the use of minimum reserve requirements on the grounds that it amounts to a tax on banks and could result in economic distortions. But all taxes introduce distortions; the real question is whether the gains outweigh the costs. The advantages… are twofold… eliminat[ing] the distortion created by providing massive subsidies to banks… gain[ing] an exceptional policy tool… while still maintaining financial stability…
Nouriel Roubini (Dec 30, 2022): More War Means More Inflation: ‘Economic and political factors will constrain… higher taxes…. Tax evasion, avoidance, and arbitrage will further complicate efforts to increase taxes on high incomes and capital (assuming such measures could even get past the lobbyists or secure buy-in from center-right parties)… increase[s] in government spending and transfers… without a commensurate increase in tax revenues. Structural budget deficits will grow even larger than they already are, potentially leading to unsustainable debt ratios…. For countries that borrow in their own currencies, the expedient option will be to allow higher inflation to reduce the real value of long-term fixed-rate nominal debt… a capital levy against savers and creditors… and it can be combined with complementary, draconian measures such as financial repression, taxes on capital, and outright default…. Because the “inflation tax” is a subtle and sneaky form of taxation that doesn’t require legislative or executive approval, it is the default path of least resistance when deficits and debts are increasingly unsustainable…
Mike Spence (Oct 12, 2022): Secular Inflation: ‘After enjoying a long period of deflationary conditions, the global economy is being pushed by a wide range of forces toward a new and more difficult equilibrium. Surges in demand that used to be accommodated by expansions in supply will now lead more frequently to higher prices…. The elasticity of global supply chains has declined…. Union organizing is expanding…. Demographic aging…. In the United States, the health sector and education account for about 20 million and 14 million jobs…. Unattractive working conditions and low compensation have persisted after the pandemic, leading to worker shortages. A new market equilibrium has not yet emerged; but when it does, it will surely include increased incomes for those working in these sectors, and hence an increase in real (inflation-adjusted) costs. More broadly, the global economy has entered a new era of frequent, severe shocks from climate change, pandemics, war, supply-chain blockages, geopolitical tensions, and other sources…
Yes, any, and all of these could happen. The world is a very strange and surprising place. But why should these risks call for more action now given the bond market's very optimistic view of the inflation outlook? It seems to me that in order to call for more action to tighten now you need to believe both (a) an overshoot and return to zero-interest-rate secular stagnatilon would not be very damaging, and (b) that the bond market is a broken gauge with respect to expectations because we have trained bond traders over the past 40 years to be optimistic about future bond prices, as it has now been a generation and a half since bond-price pessimists have been right.
But we really have warrant to believe either of things—that the bond market is a broken expectational gauge, or that secular stagnation is not destructive and damaging? And even if we have a warrant to believe one, do we have warrants to believe both?
I do not think so.
Which brings us to today’s CPI report:
BLS: Consumer Price Index Summary: ‘The Consumer Price Index for All Urban Consumers (CPI-U) rose 0.1 percent in March on a seasonally adjusted basis…. Over the last 12 months… 5.0 percent…. Shelter… by far the largest contributor… more than offset a decline in the energy index, which decreased 3.5 percent over the month…. The index for all items less food and energy rose 0.4 percent in March…. The all items less food and energy index rose 5.6 percent over the last 12 months...
Consider that there was a substantial tightening in financial conditions last August via the 10-year real interest rate channel:
That tightening is only now beginning to affect demand and spending and has not had time to impact finlation. And consider the the collapse of SVB has meant further financial tightening—of undertain magnitude, admittedly—as it has triggered a reduced willingness to lend at a constant real interest rate.
So I simply do not see the case for further Federal Reserve interest-rate tightening now.
MUST-READ: Blight as a Natural Consequence of Population Decline…
are there negative externalities from neighborhood “blight”? Almost certainly yes. Exactly how and where do they arise? Ah! That is a very hard question:
Matt Yglesias: Almost half of American counties have shrinking population: ‘state and local leaders who are worried about population decline really ought to look at the whole map and consider reading Matthew Yglesias’ seminal 2020 book “One Billion Americans.”… In a scenario of population decline, though, the market value of a house falls below the replacement value of the structures…. In years of reading about this, though, I’ve never heard of a city that has a magic formula to deal with vacant lots and blight…. In a world of sub-replacement prices, the building isn’t worthless—you can still extract income out of it by renting units to people who need housing—but it isn’t worth investing in. If something is broken, the landlord usually has a legal obligation to fix it. But the incentive is to drag feet, do the minimum, cheap out, work to rule, and generally not care about whether underlying issues have actually been resolved. Because the market price is less than the replacement cost, it’s not worth the landlord’s while to re-invest in the property. The name of the game is to just maximize the value you can extract before the thing crumbles and adds to the stock of local blight…. If overall population growth is good — and it is good — then immigration is an easy short-term path to achieving it and we should take advantage of that…
IMF: World Economic Outlook, April 2023: ‘The outlook is uncertain again amid financial sector turmoil, high inflation, ongoing effects of Russia’s invasion of Ukraine, and three years of COVID…
David Guarino: A theory of constraints!: ‘Three discrete constraints that must be tackled for these technology services to be Good: Situational awareness of problems…. Design capacity…. Technology change capacity…
Read Brad DeLong's Grasping Reality in the Substack app
Available for iOS and Android
¶s:
Doug O’Laughlin: AI Foundations Part 1: Transformers, Pre-Training and Fine-Tuning, and Scaling: ‘Foundationalpapers: Attention is All You Need (Transformer Model) December 2017… Improving Language Understanding by Generative Pre-Training(GPT-1) June 2018… BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding (BERT) October 2018… Large Language Models are Unsupervised Multitask Learners(GPT-2) February 2019… Scaling Laws for Neural Language Models (OpenAI Scaling Paper) January 2020… Language Models are Few-Shot Learners (GPT-3) May 2020… Training Compute Optimal Large Models (Chinchilla) March 2022…. The input layer is a sequence of word embeddings, and each word embedding is called a token. For each token, three linear projections are computed, the Query (Q), Key (K), and Value (V) vectors. These are created using weight matrices that are learned during the training process. There are often billions of tokens in each model. The attention score is calculated by taking a specific token's dot product of the Q vector (Query) with the K vectors of all other tokens in the sequence. This creates a score representing the importance of the other words in the sequence. The attention score holds each weight’s importance in the model…. t\The multi-head attention mechanism performs this multiple times with different weight matrices, allowing the model to focus on different aspects of the input. The self-attention mechanism allows the transformer to dynamically weigh the importance of each token in the input sequence when making predictions or generating text, better capturing long-range dependencies in the data. Multi-head attention creates a level of parallelism that didn’t exist in prior neural networks…
Tressie McMillan Cottom: Why I Keep My Eyes — and My Mind — on the South: ‘North Carolina Democratic representative Tricia Cotham switched her party affiliation to Republican…. A democratically elected representative in a district that went for Biden by 23 points in the last presidential election, Cotham gives Republicans a legislative supermajority in North Carolina with the switch. Cotham has run on a platform that champions bodily autonomy, among other things. In response to questions posed at a news conference about the impetus for her party switch, Cotham refused to name a single policy that motivated her. Nor would she say if she would maintain her position on abortion…. Downstate conservatives who break with the G.O.P. are just as likely to be censured (if differently) as Democrats who chant in the Tennessee House. The South is, above all, a political climate that rewards loyalty. It is hard to imagine turning a candidate only to allow her to be an independent…
The Fed's own staff suggested that the tightening of credit in the Senior Loan Officer's Survey may be equivalent to about 100 bps of additional increases in the Fed Funds rate. And that has barely begun to bite. No wonder the bond market is saying what it is saying. It is noteworthy, however, that the future Fed Funds rate cuts priced by the bond market likely reflect the mean/average of a distribution at each point the market expects the rate to be cut. The distribution around this mean, I strongly suspect, has a negative skew, skewed toward lower rates. In negatively skewed distributions, mean < median < mode. Inference: the modal value of the distribution expects fewer Fed Funds rate cuts than what the bond market chatter heads on TV keep saying. Meanwhile, the people you cite, quite amazingly for some of them (given their prior jobs), choose to ignore all of that evidence, which is what we really have about a future scenario (as you correctly point out). I don't know what to say about the things they are conjuring up to justify a policy action. Don Quixote at the windmills? It is one thing to give an informed recommendation, but quite another to make up word salads describing all manner of future Armageddons on which the Fed should act now. Why does this pass for policy advise? It very much reminds me of Very Serious People, who wanted fiscal austerity now just because they projected all manner of bad things happening in the future.
I'm awfully glad for a voice of sanity in the inflation debate. I'm very fearful that the Fed will not accept yes for an answer to the question, are inflation and expectations of inflation well controlled? The institutional design off the Fed insures regulatory capture by bankers.
The Fed's own staff suggested that the tightening of credit in the Senior Loan Officer's Survey may be equivalent to about 100 bps of additional increases in the Fed Funds rate. And that has barely begun to bite. No wonder the bond market is saying what it is saying. It is noteworthy, however, that the future Fed Funds rate cuts priced by the bond market likely reflect the mean/average of a distribution at each point the market expects the rate to be cut. The distribution around this mean, I strongly suspect, has a negative skew, skewed toward lower rates. In negatively skewed distributions, mean < median < mode. Inference: the modal value of the distribution expects fewer Fed Funds rate cuts than what the bond market chatter heads on TV keep saying. Meanwhile, the people you cite, quite amazingly for some of them (given their prior jobs), choose to ignore all of that evidence, which is what we really have about a future scenario (as you correctly point out). I don't know what to say about the things they are conjuring up to justify a policy action. Don Quixote at the windmills? It is one thing to give an informed recommendation, but quite another to make up word salads describing all manner of future Armageddons on which the Fed should act now. Why does this pass for policy advise? It very much reminds me of Very Serious People, who wanted fiscal austerity now just because they projected all manner of bad things happening in the future.
I'm awfully glad for a voice of sanity in the inflation debate. I'm very fearful that the Fed will not accept yes for an answer to the question, are inflation and expectations of inflation well controlled? The institutional design off the Fed insures regulatory capture by bankers.