點陣圖不是承諾，是訊號

Thank you to the Central Bank of Türkiye for the opportunity to speak to you.¹ I look forward today to hearing perspectives on the global economy and Türkiye's role in it, but I thought the most constructive use I can make of the valuable time you have granted me is to offer a brief update on the U.S. economy and then offer some thoughts on central bank communication.

感謝土耳其共和國中央銀行，給我這個機會與各位談話。今天我很期待聽聽各位對全球經濟、對土耳其在其中扮演什麼角色的看法；不過，各位撥給我的時間很寶貴，我想最有建設性的用法，是先簡短報告美國經濟的近況，再談一談中央銀行怎麼溝通。

As you know, in September the Federal Open Market Committee (FOMC) voted to raise our policy rate 25 basis points to 3.75 percent to 4 percent after nine months during which we held it steady. When monetary policy changes in this fashion, one question that most people ask is, what comes next? And I promise that I will do my best to answer that today, but first I will address another question raised by this shift in policy, which is, what changed? Was there new evidence, or did my thinking about the economy change? It turns out that the answer to this second question has a lot to do with the first one—about where policy goes from here.

各位都知道，九月時，聯邦公開市場委員會（FOMC）表決把政策利率調升 25 個基點，到 3.75% 至 4%；在那之前，我們已經按兵不動九個月。貨幣政策這樣轉向，多數人會問：接下來呢？我保證今天會盡力回答。但我想先回答這次轉向引出的另一個問題：到底變了什麼？是出現了新證據，還是我對經濟的看法變了？答案揭曉，第二個問題和第一個問題——政策接下來往哪裡走——關係密切。

I have mentioned that last month's increase in the federal funds rate came after nine months of holding it steady. When I ask "What changed?" I am obviously asking what changed in the seven weeks between July 29, when the FOMC voted to hold rates steady, and September 16, when we raised them. But I am also asking about what changed from the second half of 2025, when the FOMC reduced rates 75 basis points over three successive meetings. While it might seem to some that monetary policy has turned on a dime, or on one or two data points, in my case my decision last month was the culmination of factors that developed over the past year, and I would like to describe them.

我說過，上個月調升聯邦資金利率之前，利率已經持平九個月。我問「變了什麼」，問的當然是這七週之間變了什麼：7 月 29 日，FOMC 表決維持利率不變；9 月 16 日，我們升息。但我也在問：和 2025 年下半年相比，又變了什麼？那時 FOMC 連續三次會議，一共降息 75 個基點。有些人或許覺得，貨幣政策說轉就轉，一兩個數據就讓它掉頭。但就我而言，上個月的決定，是過去一年逐步累積的因素水到渠成。我想把這些因素說清楚。

When the FOMC cut rates from September through December 2025, we did so with inflation fairly close to our 2 percent target, after accounting for tariff effects that research found were passing through measures of inflation. At the same time, there was significant evidence of a weakening labor market, with an increase in unemployment over the first eight months of the year and very low job creation. Though the picture was less sharp after official data was interrupted because of the government shutdown that began on October 1, other data supported the view that the balance of risks to the FOMC's employment and inflation goals seemed clearly skewed toward employment. I considered our cuts to the policy rate last year as insurance against an economic slowdown.

2025 年 9 月到 12 月 FOMC 降息時，扣掉研究發現正一路傳導進各項通膨指標的關稅效應，通膨其實相當接近 2% 的目標。同一時間，勞動市場轉弱的證據很多：那年前八個月失業率上升，新增就業少之又少。10 月 1 日政府停擺，官方數據中斷，畫面變得模糊一些；但其他數據都支持同一個判斷：在 FOMC 的就業與通膨兩大目標之間，風險明顯偏向就業。我把去年那幾次降息，看作替經濟放緩投保。

But then, over the course of the first half of this year, the labor market appeared to stabilize, while progress on inflation stalled due in part to the conflict in the Middle East that drove energy prices very high. I supported no change in the policy rate in the spring and summer with the hope that the conflict would end soon and that the oil price surge would not have a lasting effect on inflation. And there were some signs of an easing of inflation. While headline inflation seesawed based on oil prices and the state of the Middle East conflict, core personal consumption expenditures (PCE) inflation—which excludes volatile food and energy prices—moderated to 0.1 percent in June and an initial estimate of 0.2 percent in July, later revised to 0.1 percent.

然而，今年上半年，勞動市場看來穩住了，通膨的進展卻停了下來，部分原因是中東衝突把能源價格推得極高。春天和夏天，我都支持利率不動，寄望衝突早日結束、油價飆漲不會在通膨上留下長久的烙印。通膨也確實露出緩和的跡象。整體通膨隨著油價、隨著中東戰況上下擺盪；但核心個人消費支出（PCE）通膨——排除波動劇烈的食品與能源——六月放緩到 0.1%，七月初值 0.2%，後來修正為 0.1%。〔譯註：皆為月增率。〕

But other forces were undermining my faith in this progress. First, hopes for a quick ending to the Middle East conflict faded, and experts warned that low inventories and damaged infrastructure could keep oil prices high through 2027. Second, evidence mounted that the artificial intelligence buildout was significantly driving up high-tech consumer prices, and projections for the size of that buildout ballooned. Third, continuing trade conflicts threatened new tariffs that could put upward pressure on inflation yet again. In most cases, these were forces that became clearer as the year passed, swamping the fleeting signs of progress toward 2 percent inflation. When the first inflation reading for August came in hot just before the FOMC's September meeting, it was impossible to deny that inflation was still too high and not making sufficient progress toward our target.

但另有幾股力量，正一點一點侵蝕我對這份進展的信心。第一，中東衝突速戰速決的希望落空，專家警告：庫存偏低、設施受損，油價可能一路高到 2027 年。第二，愈來愈多證據顯示，人工智慧的建設潮正大幅推高高科技消費品的價格，而這波建設的規模預估，也一路膨脹。第三，貿易衝突沒完沒了，新關稅的威脅可能再一次推升通膨。這些力量，多半是隨著時序推移才愈來愈清楚，淹沒了那些朝 2% 邁進、稍縱即逝的跡象。九月會議前夕，八月的第一份通膨數據燙手出爐，再也無法否認：通膨仍然太高，朝目標前進得不夠。

While some suggested that monetary policy shifted in September based largely on the single data point of consumer price index inflation for August, I hope it is now apparent that this wasn't the case for me. Instead, it was a preponderance of evidence over several months that the risks for monetary policy had shifted, reflecting a strengthened labor market and a range of persistent inflationary forces. For me, this led to the judgment that the policy setting that the FOMC maintained from December 2025 through September of this year would not be sufficient to return inflation to 2 percent in a timely manner. My decision to change the stance of policy emerged over time because it is not one that I take lightly. With evidence that economic activity is strengthening in the second half of this year, I am not greatly concerned that tighter monetary policy threatens a damaging slowdown in the economy. But I am concerned that the recent acceleration in inflation—after what soon will be five and a half years of it above the FOMC's target—will lead consumers, investors, and price-setting businesses to revise up their expectations for future inflation.

有人說，九月的貨幣政策轉向，主要是因為八月消費者物價指數（CPI）這一個數據點。我希望現在大家看得出來，對我而言並非如此。真正的原因，是好幾個月累積的優勢證據都指向同一件事：貨幣政策面對的風險已經轉移——勞動市場轉強，一連串頑強的通膨力量揮之不去。對我來說，這導向一個判斷：FOMC 從 2025 年 12 月維持到今年九月的政策設定，不足以讓通膨及時回到 2%。改變政策立場的決定，在我心中是慢慢成形的，因為這不是我能輕率做出的決定。有證據顯示，今年下半年經濟活動正在轉強，所以我不太擔心更緊的貨幣政策會讓經濟陷入傷筋動骨的放緩。我擔心的是：通膨高於 FOMC 目標，很快就滿五年半了，近來又再加速，這可能讓消費者、投資人、以及替商品定價的企業，調高他們對未來通膨的預期。

Based on some key data released last week, I see the economy in roughly the same place as it was at the time of the FOMC's September meeting. While the number of jobs created was down overall, the employment report indicated that the labor market continued to be solid and stable in September. The unemployment rate remains relatively low and near the median of policymakers' projections of its longer-run level, while payroll gains are in the range of estimates of breakeven to keep the unemployment rate steady. August inflation data, which included revisions to the government's methodology, showed monthly core PCE inflation of 0.25 percent and the 12-month change at 3 percent. Looking at the history of 12-month core inflation, it has been between roughly 2.5 percent and 3.0 percent since the spring of 2024. This is obviously higher than we want, above our target, and not showing sufficient progress. Overall, the new data reinforce my view that the labor market is stable and inflation is too high. For at least the near term, policy will be focused on the inflation side of our mandate.

根據上週公布的幾項關鍵數據，在我眼中，經濟大致還停在 FOMC 九月會議時的位置。新增就業的總數雖然下降，但就業報告顯示，九月的勞動市場依然扎實、穩定。失業率仍然偏低，接近決策者對其長期水準預估的中位數；非農就業的增幅，也落在「讓失業率持平所需」的損益兩平估計區間之內。八月的通膨數據納入了政府統計方法的修訂，核心 PCE 月增 0.25%，年增 3%。回頭看核心通膨 12 個月變動率的歷史，自 2024 年春天以來，一直在大約 2.5% 到 3.0% 之間。這顯然比我們想要的高，高於目標，而且進展不足。整體而言，新數據強化了我的看法：勞動市場穩定，通膨太高。至少在短期之內，政策的焦點會放在雙重使命中的通膨這一側。

For the remainder of my remarks, I want to discuss a particular way of communicating the expected future stance of monetary policy that has the flavor of forward guidance but isn't forward guidance. Let me illustrate what I have in mind with an example.

接下來的時間，我想談一種溝通貨幣政策未來立場的特定方式：它帶著前瞻指引的味道，卻又不是前瞻指引。我舉個例子說明。

Suppose that a majority of FOMC voters decide that policy needs to be more restrictive and that it is fairly clear to the public that some tightening is likely. What needs to be determined by both policymakers and markets is the number of rate hikes and the pace of rate hikes. Each policymaker thinks about his or her anticipated appropriate policy path to assist them in moving the economy toward the Fed's dual mandate. Markets estimate what the FOMC voters believe is the appropriate path of policy to assist them in pricing various assets.

假設 FOMC 多數有投票權的委員認定，政策需要更緊縮，而且大眾也相當清楚，一定程度的升息很可能會來。決策者和市場都得弄清楚的是：升幾次，升多快。每位決策者心中，都有一條自己認為適當的政策路徑，用來幫助經濟朝聯準會的雙重使命前進。市場則要估計 FOMC 投票委員心中認為適當的路徑，用來替各種資產定價。

So how could policymakers communicate the expected policy path to markets? There are several options. To explain, let's assume the number of hikes that policymakers have in mind is three 25 basis point hikes so that they anticipate policy will ultimately be 75 basis points higher than it currently is.

那麼，決策者要怎麼把預期的政策路徑告訴市場？有好幾種選項。為了說明，假設決策者心中想的是升息三次、每次 25 個基點，也就是預期政策利率最終會比現在高 75 個基點。

First, policymakers could choose to say nothing about the expected path of the policy rate. But that approach could surprise markets and create volatility. Without additional information, in an extreme case, they could price in zero or even potentially five hikes in future meetings. The point is that if too few or too many hikes are priced in, the change in financial conditions would mean too little or too much of an effect on economic activity.

第一，決策者可以對政策利率的預期路徑隻字不提。但這種做法可能讓市場措手不及，製造波動。沒有額外資訊，極端的情況下，市場可能在未來幾次會議反映零次升息，甚至可能反映五次。重點是：市場反映的升息太少或太多，金融情勢的變化就會對經濟活動造成太小或太大的影響。

Second, policymakers could use a strong form of forward guidance and communicate their support for raising the policy rate, let's say, every other meeting by 25 basis points until a total of 75 basis points worth of hikes has been enacted. The market would respond by pricing in 75 basis points of hikes over five meetings. While this eliminates any uncertainty about the future path of policy, it ignores the possibility that incoming data would suggest going faster or slower or hiking more or less than 75 basis points. Indeed, locking in a policy path over such a long period is unlikely to lead to good outcomes given the uncertainty around how the economy will evolve over time.

第二，決策者可以採用強形式的前瞻指引，表明他們支持每隔一次會議升息 25 個基點，直到總共升息 75 個基點為止。市場的反應會是：在五次會議之內反映 75 個基點的升息。這消除了未來政策路徑的一切不確定性，卻無視一種可能：新進數據也許顯示應該走快一點或慢一點、升多一點或少一點。事實上，經濟會怎麼演變充滿不確定，把政策路徑鎖死這麼長一段時間，不太可能帶來好結果。

Finally, there is something in between these two approaches that I will call the signaling option. What I have in mind is that policymakers could signal that the policy rate will likely be hiked 75 basis points over some time interval—say, the next six months. However, policymakers do not say what the pace of hikes will be or how large of an increment the rate hike will be. And we could emphasize that the course of monetary policy is not predetermined and will depend on incoming data and its implications for our dual-mandate goals. This approach provides some information regarding what the terminal rate will be after hiking but allows the pace and the size of rate hikes to be data dependent. Simply put, policymakers could signal where they are likely headed while acknowledging that there is no fixed final destination—except for the achievement of price stability and maximum employment.

最後，在這兩種做法之間，還有一種我稱為「訊號選項」的做法。我的意思是：決策者可以發出訊號，表示在某段期間內——比方說，未來六個月——政策利率很可能調升 75 個基點。但決策者不說升息的步調，也不說每次升多少。我們還可以強調：貨幣政策的路線並非預先注定，而是取決於新進數據，以及這些數據對雙重使命目標的意涵。這種做法對升息結束後的終點利率給出一些資訊，同時讓升息的步調和幅度依數據而定。簡單地說，決策者可以透露大概要往哪裡去，同時承認沒有固定的終點站——唯一的終點，是物價穩定與充分就業。

I see Federal Reserve policymakers' communications fitting in this last approach, both in their public remarks and their quarterly submissions of economic projections. The Summary of Economic Projections (SEP) is serving that signaling role for policymakers. The projections submitted at the September meeting showed that a large majority of FOMC participants believe it will be appropriate to raise rates at least once more this year. Sixteen of the 18 participants who submitted dots anticipated at least one more hike during the two remaining meetings this year. Four of those 16 expected two additional hikes. It gets a little complicated reading the dots for the full year of 2027, since they may reflect hikes early in the year and then cuts later in the year, but eight participants expect that rates will be 50 basis points higher at the end of 2027 than they are today.

我認為，聯準會決策者的溝通——無論是公開談話，還是每季提交的經濟預測——都屬於最後這一種。經濟預測摘要（SEP）正在替決策者扮演發出訊號的角色。九月會議提交的預測顯示，FOMC 絕大多數與會者認為，今年至少再升息一次是適當的。提交點陣圖的 18 位與會者中，有 16 位預期今年剩下的兩次會議裡，至少還會再升息一次；這 16 位裡，有 4 位預期再升兩次。2027 全年的點陣圖就比較難讀了，因為那些點可能反映年初升息、年底降息；但有 8 位與會者預期，2027 年底的利率會比現在高 50 個基點。

Between the SEP and policymakers' outlook speeches that encompass their interpretation of additional economic data, markets have interpreted that the Committee is likely to raise the policy rate 50 basis points in the coming months. According to futures market prices for the federal funds rate as of yesterday, traders see an 85 percent chance of at least one hike by the end of the FOMC's December meeting and nearly 20 percent chance of two hikes. And, by the March 2027 meeting, markets see nearly an 80 percent chance of at least two hikes and a 33 percent chance of three or more hikes. This signaling helps to anchor the path of short-term interest rates but provides flexibility in adjusting rate hikes based on incoming data.

綜合 SEP、以及決策者在展望演講中對後續數據的解讀，市場的理解是：委員會很可能在未來幾個月調升政策利率 50 個基點。根據截至昨天的聯邦資金利率期貨價格，交易員認為到 FOMC 十二月會議結束時，至少升息一次的機率是 85%，升息兩次的機率將近 20%。到 2027 年 3 月的會議，市場認為至少升息兩次的機率將近 80%，升息三次以上的機率是 33%。這樣的訊號有助於錨定短期利率的路徑，同時保有依新進數據調整升息的彈性。

If the economic data continue to come in as expected, I anticipate additional hikes to support a timelier return of inflation to our 2 percent goal. But there is some flexibility about when those hikes will occur. The hikes do not need to come at consecutive meetings, but they should be in place in an acceptable period of time.

如果經濟數據繼續如預期般出爐，我預期還會再升息，好讓通膨更及時地回到 2% 的目標。但這些升息何時發生，保有一些彈性。升息不必在連續的會議進行，但應該在可接受的時間之內到位。

I think I will stop here to make sure there is enough time for our discussion.

我想就在這裡打住，留下足夠的時間讓我們討論。