Blogs

Here’s the big picture to open the series. In our new randomized-information experiment with Turkish firms, we created clean, exogenous shifts in inflation expectations and followed those firms into their real decisions using linked admin data (credit, employment, sales, FX). Seven facts fall out. Information nudges work even when inflation is high and attention is elevated; belief changes pass powerfully into firms’ own price, wage, and cost expectations (roughly sixty percent pass-through); and when firms’ inflation expectations move down, they become more optimistic about the economy, pull back on expensive long-term borrowing in favor of short-term finance, re-tilt their currency exposures (more FX liabilities, less net FX assets), and then hire and sell more while running leaner inventories. In short: beliefs are not just survey answers—within months, they show up in balance sheets, payrolls, and shelves.

Why begin here? Because this experiment gives us causal traction on a question that matters in any macro climate, not only in high inflation: do expectations move behavior in ways policymakers and firms can use? Yes. Targeted, credible information can reduce uncertainty and steer near-term plans; belief shifts are state-dependent (they bite harder when volatility and persistence are high); and the channels we uncover—pricing, wage setting, credit maturity, FX hedging, hiring, inventories—explain why the same “news” can look expansionary in some regimes and stagflationary in others. It also reframes communication: managing the second moment (dispersion/uncertainty) can stabilize activity even when mean forecasts budge only modestly.

Expectations are the economy’s compass. In tranquil periods, people plan further ahead, disagreement is low, and professionals, firms, and households mostly point in the same direction. In transition or regime-shift episodes—say a sequence of policy surprises or a sharp disinflation—attention snaps to fresh signals, horizons shorten, and small differences in how agents read the same data get magnified. That’s when salience rules: fuel at the pump, a CPI print, an exchange-rate swing, an unusually hot month. Our research program follows those moments across audiences and outcomes, comparing calm and turbulent episodes rather than treating “the expectation channel” as a constant.

The first lesson is measurement. Instead of a single number, we elicit a compact distribution of beliefs (optimistic, most-likely, pessimistic), which lets us track both the level and the uncertainty each firm feels. When we supply short, credible public signals before elicitation, the distribution tightens. That matters because plans are set under uncertainty: compressing the range firms consider plausible lengthens horizons, reduces precaution, and makes pricing and hiring less jumpy. In the experiment, that uncertainty compression lasts beyond the initial read—long enough to affect the cadence of adjustments in wages, list prices, and contracts.

The second lesson is transmission. When aggregate inflation expectations move, firms’ own price, wage, and unit-cost expectations move almost one-for-one by comparison with low-inflation settings. That high pass-through is exactly what theory predicts when perceived inflation persistence and the frequency of price changes are elevated: firms act pre-emptively rather than waiting for costs to show up on invoices. In practice, that means belief management isn’t cosmetic—anchoring the outlook can blunt costly price-setting cascades before they start.

The third lesson is finance. Belief changes reorganize balance sheets. When a firm expects lower inflation ahead, locking in today’s long-term nominal rates looks expensive; firms delay long-term borrowing, lean more on short-term credit for working capital, and accept slightly higher near-term pricing rather than fix rates for years. Currency choices shift too: with less expected depreciation, FX liabilities look safer and FX asset hoarding less attractive, so firms re-tilt toward FX borrowing while running down net FX positions. None of this requires a banking shock—quantities and loan rates rise together when belief-driven demand is doing the work. These pivots are exactly the kind of margin policymakers and treasurers need to anticipate during disinflation or volatile energy episodes.

Finally, the real side. The same firms that revise inflation down become more optimistic about the economy and their own prospects—and then behave accordingly: sales and employment rise, while inventories fall as managers rely less on precautionary stockpiles. When expectations swing up, the mirror image appears: tighter real outlooks, more hedging, and a tilt toward short-termism. Put together, these facts explain why expectation management is a first-order stabilization tool. They also explain why “the same” announcement can pull different levers across regimes: in a low-volatility environment with anchored persistence, guidance mostly trims uncertainty; in a choppy environment with salient shocks, guidance shapes both the level of beliefs and a half-dozen concrete choices that follow.

Further reading: "Inflation Expectations and Firms' Decisions in High Inflation: Evidence from a Randomized Control Trial"  (PDF Link) (joint with Emrehan Aktug and Huzeyfe Torun) 

Monetary policy and inflation releases don’t just move markets—they move managers. In our study of Turkish firms from 2015 to 2024, we track what happens to companies’ beliefs and choices in the narrow days around central-bank meetings and monthly CPI announcements. The headline: firms pay attention, they update quickly, and they act. Rate surprises and inflation surprises shift 12-month inflation expectations, reshape optimism about the economy, and show up in day-to-day finance—especially in foreign-exchange management. These effects are stronger in choppy, high-volatility periods than in calmer times.

Each month, firms answer the Business Tendency Survey on rolling dates; some submit just before an announcement, others just after. We compare “pre” vs. “post” respondents within a tight five-day window, and we interact that timing with an orthogonalized surprise (the unexpected part of policy rates or CPI measured from up-to-the-minute Bloomberg forecasts). The design controls for firm, sector-time, and province factors, and we show the before/after timing is plausibly random with respect to firm traits. In short: clean, high-frequency identification of how news shifts beliefs.

State dependence is the rule: In relatively stable years, a surprise tightening looks “textbook”: firms become more pessimistic about activity and lower their inflation expectations—much like professionals’ demand-side reading of policy. In turbulent, high-inflation years, the narrative flips: firms raise inflation expectations after a surprise hike, and they still turn more pessimistic. That’s the cost-push reading—higher rates mean higher financing costs, which managers expect to pass through to prices even as demand softens. The same state dependence shows up after CPI news: positive inflation surprises lift expected inflation in all regimes, but the pass-through is markedly larger when volatility is high.

From beliefs to prices, costs, and plans: These belief shifts aren’t cosmetic. After policy surprises, firms revise not only aggregate CPI/PPI expectations but also their own price and unit-cost expectations, and they trim sales and employment outlooks. Smaller and highly-leveraged firms are more sensitive on almost every margin, consistent with tighter financing constraints and higher exposure to borrowing-cost risk. In addition, with firm-level daily transaction data, we see managers reposition currency exposures within days. After a positive inflation surprise, firms buy more FX and sell less—net FX rises—a hedge against perceived depreciation. After an unexpected rate hike, firms reduce net FX positions (more sales, fewer purchases), consistent with an appreciation narrative dominating other channels in the immediate window. These responses are sharpest for large and lower-leverage firms.

Why this matters beyond high inflation? The study underscores two general lessons. First, communication works through levels and second moments: tightening belief dispersion and clarifying the near-term path can stabilize plans even if average forecasts move modestly. Second, transmission is regime-dependent. The same announcement can look demand-driven in calm times and cost-push in turbulent ones—so the tone, timing, and targeting of messages should adapt to the state of the world.

To conclude, firms are not passive recipients of macro news. In a tight, well-identified window around policy and CPI releases, they update beliefs, rethink pricing and costs, and re-hedge FX—fast. That makes clear, state-aware communication a first-order policy tool and a practical playbook item for CFOs navigating uncertain cycles.

Further reading: "How Do Monetary Policy and Inflation Announcements Affect Firm Expectations in High Inflation?"  (PDF Link)  (joint with  Emrehan Aktug)

Oil is the most conspicuous “macro signal” most people see. When the price at the pump jumps, households feel it in their wallets, managers see it in their cost sheets, and market forecasters fold it into their models. Across three studies using Turkish microdata, a coherent picture emerges: oil shocks don’t just raise measured inflation; they quickly reshape inflation expectations—and those belief shifts spill into real decisions about pricing, hiring, and finance.

(1) Start with the cross-section of decision-makers. Using harmonized surveys for households (CTS), firms (BTS), and professional forecasters (SMP), we track how a simple, salient shock—fuel prices—ripples through beliefs. A 1% rise in fuel prices lifts households’ one-year-ahead inflation expectations by about 0.35%, with stronger effects for lower-income respondents (0.37% vs 0.28% for higher-income). Professionals raise 12-month expectations by 0.36%, but longer-run beliefs are more anchored (around 0.18% at five years), and within the professional sample, financial-sector forecasters extrapolate more than real-sector peers. Firms sit in the middle: a 1% fuel increase pushes their one-year CPI expectations up by ~0.38%, with larger effects in energy-intensive industries. These are not sterile survey moves: firms report weaker sales and macro outlook, and we see real adjustments—sales, purchases, and employment edge down, while supplier networks widen as managers diversify risk. The common reading is stagflationary: higher expected inflation paired with a softer real outlook.

(2) Next, zoom out from the gas station to global supply news. Around OPEC announcements, we isolate “oil-supply news” shocks that move futures prices independently of local conditions and ask what firms do with that information. The answer is: quite a lot, for quite a while. At the aggregate level, a one-standard-deviation positive oil-supply shock raises realized core CPI and firms’ average CPI expectations by roughly 3 percentage points at the peak, with effects that persist for up to 15 months (peaking around month ~10). At the micro level, the same shocks push up firms’ CPI/PPI beliefs and their own price and unit-cost plans, while capacity utilization and business outlook fall. Heterogeneity is telling: smaller and more leveraged firms react almost twice as much as larger, better-capitalized peers, consistent with financial constraints amplifying cost-push news. And because Türkiye did not operate a binding ETS during the sample, carbon-price news shows no meaningful impact on firms’ beliefs or activity—an institutional null that helps attribute the action to oil rather than policy-driven carbon costs.

(3) Finally, look at the highest-frequency link from beliefs to finance. Exploiting city-level pump-price changes within a five-day window, we find a rapid pass-through from oil to firms’ expectations: a 1% increase in local fuel prices raises one-year-ahead firm inflation expectations by ~0.3 percentage points within days—about a 30% pass-through. That belief shock lines up with a clean maturity pivot in corporate borrowing: short-term credit (working capital) rises by roughly 0.5%, while long-term borrowing (investment) falls by about 0.5%. Quantities and rates move together, pointing to demand-driven reallocation rather than a contraction in bank supply. In other words, salient energy costs compress horizons: managers bridge cash-flow pressure now and delay horizon-lengthening projects until uncertainty clears.

Put together, these facts map a tight chain from oil → expectations → choices. Households trim durable plans and darken their macro outlook; firms revise CPI and cost beliefs up, mark down sales and utilization, and rotate toward short-term finance; professionals adjust near-term forecasts while keeping longer horizons relatively anchored. When shocks are large or persistent, those micro decisions scale into network effects: in firm-to-firm data, oil supply shocks shrink sales, purchases, and the number of trading partners, especially for financially constrained firms—evidence that energy shocks propagate through production linkages, not just price indexes.

Why this matters beyond a single episode or country is straightforward. In energy-importing economies, oil is both an input and an information beacon. Because it is visible and frequent, it gets overweighted in belief formation—especially when macro volatility is high. That makes expectations state-dependent: the same policy or data news will move beliefs more when energy is noisy, and balance sheets will mediate how those beliefs become action. The policy playbook follows: communicate clearly about the persistence of energy shocks; separate near-term cost-push from demand pressure; and pair guidance with liquidity tools that keep short-term working-capital channels open without turbocharging long-run risk. Stabilizing the second moment—reducing uncertainty and disagreement—can be as valuable as nudging the mean forecast, because it lengthens planning horizons and tempers the rush to precaution.

Further reading: (1) "Do Fuel Pump Prices Shape Expectations? Evidence From Households, Firms and Market Participants" (PDF Link)  (joint with Velihan Başpınar) 

(2) "How Do Oil Supply and Carbon Policy Affect Firms’ Expectations and Decisions?"  (PDF Link) (joint with Emrehan Aktug, Huzeyfe Torun and  Cihan Yalçın)  

(3) "Oil Price Shocks: Firms' Expectations and Borrowing Decisions"  (joint with Emrehan Aktug and Velihan Başpınar)  

When the weather is “off,” the economy feels it. Linking Türkiye’s household and firm surveys to province-day meteorology, we track what locally unusual months do to prices, expectations, real activity, and credit. A one–standard-deviation temperature anomaly raises near-term food CPI for several months—peaking around months 4–6—and then mean-reverts. That salient price impulse shows up in beliefs fast: households’ 12-month CPI expectations rise by ~1.5%, and food-sector firms’ expectations rise by ~0.6%, alongside higher unit-cost expectations. At the same time, sentiment turns down: households report weaker income prospects, darker macro outlooks, less willingness to buy durables, and more unemployment concern—especially in rural, female, and lower-income groups. Managers in food manufacturing mark down sales expectations and overall business conditions, with SMEs and high-leverage firms reacting most. In short: transient but salient climate shocks lift expected inflation while dimming the real outlook.

These belief shifts become concrete choices. On the real side, food-sector firms see monthly contractions: domestic sales (–0.8%), exports (–1.2%), purchases (–0.6%), and employment (–0.4%), with bigger drops for smaller and financially tighter firms. On the finance side, credit doesn’t just change in size—it rearranges in time: total outstanding credit rises (+1.7%) but via short-term balances (+3.6%) as long-term balances fall (–1.0%). That is classic maturity shortening: managers bridge cash-flow risk now and delay horizon-lengthening projects until uncertainty clears. The mechanism is intuitive. Unusual heat (or cold) strains yields and quality, raises cooling/heating and logistics costs, tightens working-capital cycles, and makes essential prices more salient. Beliefs move, plans compress, and firms pivot to short maturities—especially those with thinner buffers.

Why it matters for policy? Climate anomalies temporarily steepen the monetary-policy trade-off. They raise expected inflation while softening activity, so communication should separate transitory supply impulses from persistent demand pressure and speak directly to salient prices (food/energy) that anchor household and firm updating. In practice, climate-aware messaging plus targeted liquidity tools that keep short-term working-capital channels open—without fueling long-run risk—can smooth the near-term adjustment. The broader lesson is general: in any macro regime, salience and state-dependence govern expectations. When the weather is weird, beliefs bend—and the economy adjusts along prices, plans, and the maturity of credit.

Further reading: "Weather Anomalies, Climate Risk, and Inflation Expectations of Households and Firms" (PDF Link)  (joint with Y.Emre Akgündüz and Seyit M. Cilasun)

Right after the May-2024 Business Tendency Survey, we randomly showed firms one of several short, credible signals—professionals’ 12-month CPI forecast, the CBRT’s current-year or next-year projection, or the 5% medium-term target—before eliciting beliefs as a min–mode–max distribution. Treating those three points as a compact posterior (triangular) lets us read both the level of expected inflation and the uncertainty firms feel about that view.

Three results stand out. First, targeted guidance tightens belief distributions and lowers perceived forecasting difficulty; firms feel surer about the near-term path. Second, means move modestly toward the signal, with the largest pull when the message matches the decision horizon (12-month or current-year); longer-horizon messages nudge less. Third, the improvement is sticky but not permanent—it lasts roughly two to three months and then fades as new information arrives, a useful clock for how often guidance should be refreshed.

Why it matters is simple: uncertainty isn’t just vibes—it changes behavior. Higher perceived inflation uncertainty goes hand-in-hand with defensive pricing, shorter credit maturities, and delayed hiring or investment. By managing the second moment of beliefs—not only the average forecast—communication can steady plans without touching policy rates today. The practical playbook: track easy KPIs (cross-firm disagreement, “forecasting is hard” responses), refresh guidance on a regular cadence around key data and meetings, and align messages with firms’ planning horizons so the uncertainty compression is strongest where choices are being made.

Further reading: "Communication as Policy and Firm Uncertainty: Evidence from Randomized Control Trial" (PDF Link) (joint with Emrehan Aktug, Fatih Karahan and Huzeyfe Torun)

On one September day in 2021, Türkiye delivered the kind of surprise economists dream of: an unexpected policy rate cut. Because firms happened to answer the monthly survey just before or just after the shock, we can watch beliefs split in real time. “Updaters” raised their 12-month inflation expectations sharply; “non-updaters” didn’t. That gap persisted for months (then faded as other news arrived) and became a clean test of what higher expected inflation does.

It did a lot. Updaters marked up not only aggregate CPI/PPI views but also their own price and unit-cost plans, turned more pessimistic about demand, and scaled back production intent—capacity utilization fell within weeks. Finance reacted fastest: firms borrowed more and paid higher rates (a demand-driven grab for working capital, not a bank squeeze), shortened maturities, and lifted net FX as a hedge; some shifted sales toward exports to earn foreign currency. On the real side, a 10-pp rise in expected inflation was followed by small but meaningful declines in domestic sales and employment, while input purchases and wages ticked up as managers front-loaded costs.

The takeaway is general, not just “high-inflation-Turkey” specific: expectations are first-order drivers of near-term choices. A single surprise can push beliefs, and beliefs quickly rewire pricing, hiring, inventories, and balance sheets. That’s why credibility, timing, and message matter. Clear, state-aware communication that separates transitory cost impulses from persistent pressure—and liquidity tools that keep short-term working-capital channels open without stoking long-run risk—can soften the real-side fallout while beliefs settle.

Further reading: "Quasi-Experiment in Monetary Policy:  The Impact of 2021 Rate Cut on Inflation Expectations and Firm Behavior"  (PDF Link)  (joint with  Emrehan Aktug, Altan Aldan and Ünal Seven )