https://w w w ./economics/monetary-policy-conference-and-inflation-past-present-and-未来/5 月 6 日，这一天的开幕式和胡佛货币政策一样。 发表 1天前 上 2022年5月24日 本文最初由《暴躁不安》《经济学家》发表 5 月 6 日，这 一场会议最后的 回归和胡佛货币政策一样。唉，视频成绩单和所有这些都没有准备好。 ，我为我的简短演讲写了一篇论文；多亏了胡佛团队，我还得到了演讲的抄录。标题是现在和未来的热议：热议、热议和未来”。但是，但它的发布了其他小说、 物价水平理论博客很重要的零碎联想。 漂亮的重复，越来越多的包装和最近的屏幕。这是一个个性化的文章，更短它不同的尺寸： 过去、现在和未来的紧急情况：财政、冲击和财政危机 现在，我们已经出现了，很不错的 反应。 为什么？通货膨胀从何而来，是第一个问题，查理·普洛瑟（Charlie Plosser）在他对本届会议的精彩序言中给出了答案：政府基本上做了一个财政直升飞机，以特别有力的方式发送了五到六万亿美元的资金。他们寄给人们支票，一半是新的储备，一半是借来的。这是一次财政直升机投降。想象一下，这只是 6 万亿美元的公开市场操作。好吧，正如拉里刚刚告诉我们的那样，增加 6 万亿美元的 10 美元钞票和减少 6 万亿美元的100 美元钞票不会有太大的不同。如果没有赤字，肯定不会产生如此巨大的影响。 这种冲动也不是利率政策的错。利率刚刚持平。人们可以责怪美联储促成了巨大的直升机下跌，但不能归咎于巨大的利率冲击。 这就是通货膨胀的冲动，但是现在通货膨胀了吗？现在，注意力转向美联储。正如您在第一张图中看到的那样，当通货膨胀从财政冲击中消失时， 利率保持不变。 所以下一个问题是，这反应慢吗？这段时期的名义利率远低于通胀，构成额外的货币刺激，这会自行造成额外的通胀吗？还是我们只是在等待财政（或供应，如果必须的话）冲击结束？ 按照历史标准，美联储肯定落后于曲线。在这张图表中，箭头指向自 1960 年以来的每一次紧缩。在每一次紧缩中，美联储在当前通货膨胀的情况下，利率大致为一对一或多。即使在可怕的 1970 年代，美联储也从未等待一整年才采取任何行动。 为什么美联储现在花了这么长时间才采取行动？这个问题有很多答案，我会的不要深入研究它。其中一个功能可能用于病房指导，前一个小组提到了这一点。美联储保持低利率是因为美联储表示将保持低利率。有一个实验室正式的新战略说我们将保持低利率。我认为，这种策略是一条针对通货紧缩而精心打造的美丽马其诺防线，但与原来的策略一样，它忘记了：如果德国人转而通过阿登山脉会怎样？我们发现了拉里和约翰·泰勒之间的半途而废。前瞻指引的问题是，有人相信美联储会在事后采取行动吗？（时间一致性。）像我这样说不的批评者我认为被证明是错误的。美联储事后对它所承诺的事情进行了充分的保证。不幸的是，它承诺了一些在当时情况下不合适的东西，这就是指导的全部要点和难题。还有什么约翰会说，美联储应该做出的承诺包括，如果发生严重的通货膨胀，通货膨胀，那么我们会做其他事情。 美联储总体上做得太多“这是我们认为会发生的事情，所以这就是我们将要做的事情”，而应急计划太少以防万一事情不顺利它的项目。我们刚刚了解到，这可以在很大程度上以惊人的速度发生。军事规划者知道这一点。他们对他们的预测进行红队，对不太可能发生的意外事件着迷，并通过详细的自我批评调查来跟踪失败。美联储也应该这样做。 但就像震惊一样，无所作为的一年已经过去。相关的问题是我们将来会去哪里？这让我们回到了这个问题：停留这么久本身是否是额外的刺激，会导致通货膨胀？或者，假设财政冲击已经结束，在下一次冲击到来之前，通胀会自行消失吗？ 我在这里绘制了美联储在 3 月 15 日会议上对失业率、联邦基金利率和通胀的预测。 美联储认为，在没有高实际利率时期的情况下，通胀将自行发展。 我认为，这是我们大多数评论员不同意的中心前提。到m，在实际利率大幅偏低或为负的时期将加速通货膨胀。他们会说，这些预测是疯狂的。但他们真的疯了吗？ The markets, by the way, seem to agree with the Fed. Until the Fed started saying it’s going to move, markets also seemed to think inflation would go away all on its own. Well, one thing John emphasizes is to write down models. So I wrote down a model, so simple it can show up in the bottom of a slide. It says that output is lower when the real interest rate is higher, and a Phillips curve in which inflation depends on the output gap. Crucially, this model uses adaptive expectations in the Phillips and IS curves, as shown by the arrows. Then I asked the question, let us start at last year’s inflation, where the Fed projections start; assume the shocks are over, but we inherit this inflation. What happens if the Fed follows the blue line, the federal funds rate? The initial inflation is given. The blue line is given. From the model, I calculate where inflation and unemployment go from there. The dashed lines are the Fed’s projections. I think this graph encapsulates the view the Fed is way behind the curve; that low interest rates below inflation constitute additional stimulus. Inflation spirals off, at least until it gives in, raises rates sharply and causes a recession, as Larry Summers warns. That looks pretty awful. And it looks like the Fed is completely wrong in its forecasts of what will happen if it follows this fund rate path. But what if expectations are rational? Here’s a little modification of the model with the arrows pointing to the changes. What if, the real interest rate and the Phillips curve are centered at expected future inflation rather than lagged inflation? Same simulation: Put in the federal funds rate path, anchor inflation at last year’s inflation. Turn off all shocks. What happens? I obtain almost exactly what the Federal Reserve is projecting. Intuitively, the rational expectations Phillips curve looks at inflation relative to future inflation. Unemployment is low, as it is today, when inflation is high relative to future inflation. Inflation high relative to future inflation means that inflation is declining. And that’s exactly what this projection says. Rational expectations means you solve models from future to present. If people thought inflation was really going to be high in the future, inflation would already be high today. The fact that it was only five percent tells us that it’s going to decline and go away. To be clear I don’t think the Fed thinks this way procedurally. They have a gut instinct about inflation dynamics, informed by lots of VAR forecasts and model simulations. But this theory gives a pretty good as-if description, a slightly more micro founded model that makes sense of those intuitive beliefs. So the Federal Reserve’s projections are not nuts! Inflation might just go away on its own! There is a model that describes the projections. And this is a perfectly standard model: it’s the new-Keynesian model that has been in the equations if not the prose of every academic and central bank research paper for 30 years. So, it is not completely nuts. Our job is to think about these two models and think about which one is right about the world. To put the question in another way, let’s find what interest rate it takes to produce the Fed’s inflation projection. What should the Fed be doing? With adaptive expectations, to make inflation fade away as the Fed thinks will happen, we need the interest rate to be nine percent. Right now. Why? Because you need a high real interest rate measured relative to lagged inflation in order to bring that inflation down. And we’re way, way off the curve. I think this calculation encapsulates a lot of the Taylor rule view. But what if inflation expectations are rational? Again I find what interest rate path, it takes to produce the Fed’s inflation path, Here, the Fed is a bit ahead of the curve. A quick summary: The Fed’s projections are consistent with a standard New Keynesian model. This is not a nutty model. It’s been around since 1990. It’s the basis of essentially all academic macroeconomic theory. It may be wrong or right. The point of models is to help us understand what ingredients get to different views, and thereby also use evidence from other episodes to guide this one. I’m not advocating for against it. I’m just saying this is what we need to talk about. Here are the core questions.
In my view, the right answer is probably halfway in between. In the long run, people catch on. Why don’t we see the huge spiral? Because once we get to spiraling inflation, people catch on and get more rational in their expectations. In the short run, with a new event, there is likely some adaptive dyanmics. So don’t count on either one being perfectly right. But don’t count on either one being incredibly reliable either. Another, more uncomfortable way of putting the question,
美联储的预测称，稳定。如果我们不理会利率，最终通胀会平息下来。在此过程中可能会有很多短期动态，这些模型没有捕捉到这些动态。但它最终安定下来。这回答了拉里的最后一个问题。这不一定是一个错误。钍这是一个世界模型，在这个模型中，名义事物确实控制名义事物，名义利率最终会拖累通货膨胀。k% 利率规则是可能的。不一定是最佳的，但可能。 第三，有一个重要的定量问题。
美联储的预测表明通胀会很快下降。隐含地，价格非常灵活。也许不吧。 You may still think the Fed’s view is nutty, but there is now some important evidence on the table. That experience may have shifted the Fed’s view of inflation dynamics. The model of spiraling inflation and deflation also predicts that when interest rates are stuck at zero for 10 or 20 years, inflation or deflation will spiral away. And inflation didn’t do that. The Fed funds rate was stuck at zero in the US for nearly 10 years. Inflation just batted around with a stuck interest rate. In Europe, inflation was stable with a stuck interest rate for a longer period. And in Japan, inflation was stable with a stuck interest rate for 25 years. The deflation spiral never broke out. Japan lived 25 years of the Friedman optimum quantity of money. This experience may be is something that the Fed is digesting in its verbal way, and that is what the rational expectations model says. So it’s not as nuts. Next, let’s think about how inflation is going to proceed in the future. In the Fed’s projections inflation comes down very quickly. To match that I had to take a price stickiness parameter of 0.5. That means a very steep Phillips curve. One percent output gap means one percent inflation. Most people think of the Phillips curve as having been quite flat. In the 2010s, unemployment moved a lot with little change in inflation. Today inflation is moving a lot with little change in unemployment. What is the real amount of price stickiness? Or is the whole Phillips curve a mess? To get at this, and other issues, here is the response to a fiscal shock in the full equations of the standard new-Keynesian model. I add long-term debt, and I perform slightly unusual simulations: I ask how the model responds to a fiscal shock with no change in interest rate, and I ask how it responds to interest rate changes with no change in fiscal policy. This is now a full shock simulation: Rather than take initial inflation as given, I hit the model with a 1% fiscal shock, no change in interest rate, and derive the entire inflation path. What happens? Inflation rises, and then does go away on its own. This is the stanford three equation model. So it really is not nutty that inflation would go away on its own. But it takes years for inflation to go away. I cut the price stickiness parameter in half, but I still have a pretty steep Phillips curve. Using the full model adds persistence. So the Fed may also be wrong in thinking that inflation will disappear on its own quite as quickly, even if the fiscal shock has ended (as it has in the simulation) and even given its view that inflation is stable in the long run. A one-time fiscal shock can lead to drawn out inflation, not a one-time price-level jump. We may have a ways to go. In this model the 8% inflation we have just seen is only 40% of the eventual price-level rise. Now, what can the Fed do about inflation? Suppose the Fed does start raising interest rates aggressively? Here I feed the model an AR(1) interest rate, with no fiscal policy shock, to evaluate the independent effect of monetary policy. Many models implicitly specify that when the Fed raises interest rates sthere’s a big increase in surpluses. We’re just going to leave fiscal policy alone and see what happens in what is the apparently Fed’s model, if the Fed raises interest rates. The interest rate rise does lowers inflation, but only in the short run. There’s a form of Tom Sargent’s “unpleasant arithmetic”at work. Lowering inflation today raises inflation in the future. We had a fiscal shock. That has to come out of the pockets of bondholders, by inflating away bonds. The Fed can choose to inflate that away now or inflate it away in the future. But the Fed cannot get rid of the fact that there has been a fiscal shock which inflates away debt. So it has a limited power. It can smooth inflation but cannot completely get rid of inflation. What if there is a 1% fiscal shock, but the Fed responds with something like a Taylor rule? We basically add the last two figures. We get lower inflation in the short run, but much more persistent inflation in the long run. Smoothing inflation is a great thing in this model. It reduces output volatility. (Output is related to inflation relative to future inflation, so has almost no movement here.) We have one more reason for a Taylor rule: It does not give us stability, as it does in adaptive expectations models, it does not give us determinacy as it does in new-Keynesian models. It gives us quiet, the absence of volatility, which is perhaps the best of all. Now let’s think about the further future. This is the CBO’s debt and deficit graph to remind you of our fiscal situation. We don’t have just a one-time $6 trillion helicopter drop, we have an ongoing fiscal problem. The CBO’s projection is debt if nothing bad happens. But every 10 years we have a shock, and the debt goes up again like a cliff.My little black line there is a guess of what happens with occasional shocks. In sum, we have five percent of GDP structural deficits, once per decade stimulus/bailouts, and here comes Social Security and Medicare. If there isn’t more bad fiscal news, there are fiscal constraints of monetary policy. In 1980, there was only 25% debt to GDP ratio. Now it’s 100% and rising. We need coordinated fiscal and monetary policy to get out of a serious inflation. But the fiscal constraints on monetary policy are going to be four times harder this time. How? First, there are interest costs on the debt. Suppose the Fed raises interest rates to five percent. We’ve been talking about eight, nine, ten percent, so five percent isn’t that much. But today five percent interest rate is five percent of GDP debt costs, a trillion bucks. If the Fed wants to raise real interest rates 5% to fight inflation, we need a fiscal contraction of a trillion dollars a year to pay those debt costs, or else raising interest rates does not stop the inflation. Second, once inflation gets baked in to bond prices, then if we disinflate, the government has to pay bondholders a big windfall, by repaying debt in more valuable dollars. Ten percent disinflation and 100% debt to GDP ratio means a ten percent of GDP present from taxpayers to bondholders. If we don’t pay that, the monetary tightening will fail. And we’ve seen that failure in Latin America. Countries get into fiscal problems, they have inflation, they raise interest rates, that just raises debt costs, and the inflation spirals on up. 这是该效果的模拟。在这里，我采用了一个完全标准的新凯恩斯模型——没有任何关于价格水平的财政理论。我问如果美联储按照虚线加息会发生什么？在左上角的图表中，您看到利率上升而通胀下降。利率上升会降低通货膨胀。但是，较高的名义利率和较低的通货膨胀意味着较高的偿债成本和对债券持有人的礼物。新凯恩斯主义模型认为国会和政府“被动地”提供无论如何都需要财政支持，但需要多少？标题计算：这种通货紧缩需要 GDP 财政盈余的 3.5%。 如果国会对美联储说：“你疯了。我们不会那样做。” 好吧，那它不可能发生。 在第二个模拟中，我遵循相同的利率路径，但在新凯恩斯模型中，通过改变货币政策冲击的时间序列模式，您可以在较少需要财政政策的情况下获得相同的观察利率。所以在这里，我有相同的利率路径，但只有 2.3% 的 GDP 财政盈余来自国会。但你从中得到的通货紧缩减少了。在右下角，国会只提供了 GDP 财政盈余的 0.91%。利率上升，但现在通胀根本不动。仍然有盈余，因为我们仍然需要支付更高的实际利息成本。在右下角的情况下，较高的利率会提高通货膨胀率。为什么？因为在这种情况下，我们有赤字。 没有协调的财政紧缩政策，提高利率不会降低通货膨胀。即使在完全库存的新凯恩斯模型中。 1980年呢？1980 年不仅仅是货币政策。1980 年是联合货币、财政和微观经济放松管制的改革。它引发了增长的繁荣，从而引发了税收的繁荣。这正是财政改革和紧缩。我在这里绘制了 1980 年之后的基本盈余。大多数“里根赤字”只是巨额债务的更高利息成本。到 1990 年代，您可以看到财政盈余激增。花了一段时间，但伴随着货币改革而来的是财政和放松管制改革（1982 年、1986 年）。无论是通过这些改革还是仅仅靠运气，增长都起飞了，还清了很多债务。 你甚至可以在债务与 GDP 的比率中看到这一点，其中包括债务的利息成本。到 1990 年代，这些改革正在偿还债务。 假设您在 1980 年以 15% 的收益率购买债券。到 1990 年代后期，您以 4% 到 5% 的通货膨胀率买回了这些债券。你得到了每年百分之十的实际回报率，这是纳税人提供的。由于纳税人的恩惠，许多债务以更高的利率滚动。该图还警告我们，债券收益率从来都不能很好地预测通胀。 The joint fiscal and monetary policy was absent in many Latin American countries. They raise interest rates, the deficits don’t get cured, and the whole thing falls apart. That could have happened in the US as well. Last, I offer some good news. We’re forgetting a lot of the economics of the 1970s and 1980s. We can have painless disinflations. We keep talking as if we have to redo 1980, and endure a horrible recession. There is a way out. And the way out is to coordinate the fiscal, monetary, and micro economic reform to a durable new regime, move that expected inflation term in the Phillips curve, by having people convinced that it really is going down. Then we get a miraculous disinflation. It can happen. It does happen. 我举三个例子。首先，请记住 1990 年代初的通胀目标改革。这个是新西兰。GST 符号标志着商品和服务税的上升。还有一套微观经济放松管制改革。通胀目标是一项联合财政、微观经济和货币改革。通货膨胀像石头一样下降，根本没有衰退。央行从来不需要利用其独立性来重演美国 1980 年。事实证明，这不是通胀目标的重点！ 加拿大在 1990 年代初也做了同样的事情。通胀目标不只是“亲爱的加拿大储备银行，请更加严厉”。这也是财政部的一项承诺，“如果有必要，我们将提高税收以偿还我们的债务，而且只有这个通货膨胀的价值。” 它还包括微观经济改革。通货膨胀逐渐消退，就像加拿大 7 月中旬的 雪一样。 当然，正如汤姆萨金特告诉我们的，记住恶性通货膨胀的著名结局。即使利率下降，即使印制更多货币，即使短期赤字增加，通货膨胀也会停止，因为各国解决了结构性财政问题。 如果我们到了需要去通胀的地步，请记住，所有成功的去通胀，包括 1980 年的美国，都是联合货币、财政和微观经济。但他们必须改革为可信、持久、时间一致的政权。你无法为摆脱通货膨胀提供指导。你不能只发表更多关于“锚定”的演讲。 还记得“ W IN”（W hip Inflation Now w）按钮吗？当我们重播1970 年代时，它们可能会回来。当然，它们是对病房指导或期望管理的最终尝试，但没有奏效。你需要对政权进行持久的改变，以便人们了解根本的财政问题得到了解决。 但回到厄运和阴霾。我们将面临新的财政冲击，迟早会发生不好的事情，而且我们的财政空间很小。如果（何时？）中国入侵台湾，我们将面临巨大的金融和经济冲击，以及一场恶战。政府将尝试借入或印制另外 5 至 10 万亿美元来纾困、刺激、保障和打一场战争。最重要的是，比如说 150% 的 GDP 债务和未改革的权利。看到w会很有趣帽子发生了。 在我的阅读中，我们已经越过了卢比孔河。我们发现人们不再想要我们的债务，开始花钱，并导致通货膨胀。我们现在是否处于财政极限，无法再次实施大规模的赤字融资刺激计划？或者人们，人们是否认为未来的赤字将通过未来更多的盈余来偿还，并且他们会很高兴地在没有通货膨胀 的情况下借出这些万亿美元？ 所以这里是未知的理论问题编号 t w o。
我认为拉里萨默斯的愿景，优美的表达，是一个流动的限制。赤字乘以 1.5 不应大于 GDP 缺口。如果是这样，你就会得到通货膨胀。只要赤字流量低于此值，就不要担心增加债务；我们总是可以迟早还清的 。 我在这里表达的观点是，观点的看法。我们是从现在开始说的，当时是对政府未来的债权可能是当时的那样。可能的变化，也会给我们带来财政问题。 我们有很好的基本问题。让我们讨论答案。 |
|
|
来自： 素心馆801 > 《inflation》
