Banque Nationale De Belgique Sa · EURONEXT:BNB
Banque Nationale De Belgique Sa — Technical Presentation (December 2024)
37 slides from Banque Nationale De Belgique Sa's technical presentation, Dec 3, 2024
Slide 1: Decoding Equity Market Reactions to Macroeconomic News
Decoding Equity Market Reactions to Macroeconomic News Michele Modugno Berardino Palazzo Federal Reserve Board* Federal Reserve Board* National Bank of Belgium December 03, 2024 *The views expressed in this presentation are solely our responsibility and should not be interpreted as reflecting the vi
Slide 2: In this paper
In this paper We show that the reaction of the equity market to macro news is rationalized by their propagation into the real economy. Following a stream favorable macroeconomic surprises: ▶ firms revenues, profitability, and investments increase ▶ macroeconomic aggregates expand We embody the macro
Slide 3: Macro News and Asset Prices
Macro News and Asset Prices Information about the state of the economy is released daily, and market participants forms expectations. 3 / 37
Slide 4: Macro News and Asset Prices
Macro News and Asset Prices When releases differ from expectations, market participants adjust their investment strategies and asset prices change. 4 / 37
Slide 5: Macro News and Stock Prices: What We Know
Macro News and Stock Prices: What We Know However, the existence of a linear causal relation between macro news and stock prices has been deemed as controversial: ▶ Earlier studies found insignificant reactions of stock prices to macro surprises. Cutler, Poterba and Summers (1988), Hardouvelis (1987
Slide 6: A common explanation
A common explanation Macro news transmit to stock prices trough different channels: ▶ future path of the interest rate, it+j (- sometimes) ▶ future path of the risk premium, rpt+j (+?) ▶ future path of dividends, Dt+j (+) ∞ X Dt+j Pt = E j=0 1 + it+j + rpt+j The strength of each channel changes at d
Slide 7: First Finding
First Finding Activity and price news explain 34% of stock price quarterly fluctuations across the whole business cycle, with opposite signs. We exploit the entire universe of macro news: ▶ Statistical reports contain more info than the headline. e.g. the employment report contains 72 data releases:
Slide 8: Second Finding
Second Finding A stream of macro surprises are followed by both mechanical and discretionary firm’s responses. ▶ A quarter positive activity news is followed by an increase of sales, liquidity, and profitability, but also financing, and investment. ▶ On the contrary, a sequence of positive price new
Slide 9: Third Finding
Third Finding Firm-level outcomes have macroeconomic consequences A sequence of positive activity (price) news over a quarter is followed by ▶ changes in aggregate temporary production factors and, therefore, production. Hours worked, capacity utilization, and industrial production ↑ (↓). ▶ but also
Slide 10: Roadmap
Roadmap 1. Macroeconomic News 2. Activity and Price News Indexes 3. Stock Returns reaction to Macroeconomic News: ▶ Daily Frequency ▶ Quarterly Frequency 4. Firm-level responses 5. Macro aggregate responses 10 / 37
Slide 11: Macro News
Macro News Defined as: (Ai,t − Mi,t ) si,t = std(Ai,t − Mi,t ) where ▶ the actual release of a macro variable i, Ai,t ▶ the market expectations for that same release, Mi,t ▶ standardized for their historical standard deviation The data are available in the Bloomberg Economic Calendar (ECO). Mi,t is
Slide 12: Activity-related Macro News
Activity-related Macro News 12 / 37
Slide 13: Price-related Macro News
Price-related Macro News 13 / 37
Slide 14: Macro News Indexes
Macro News Indexes Differently from previous studies, we do not focus on the isolated impact of macro news. Instead, we aggregate news relative to real activity and news relative to prices in two separate indexes. na np X X anit = si,t wi pnit = si,t wi i=1 i=1 where na and np are the number of real
Slide 15: Macro News and Daily Stock Returns
Macro News and Daily Stock Returns Our first set of results is based on the following regression: rxt = α + βanit + γpnit + δmpst + ϕrxt−1 + ut . where rxt = log(S&P 500t ) − log(S&P 500t−1 ) and mpst is Bu, Rogers and Wu (2021) monetary policy shock. anit pnit mpst rxt−1 R2 2 Radj 0.07 0.01 -0.06 -
Slide 16: Quarterly Aggregation
Quarterly Aggregation Altavilla et al. (2017) show that 45% of bond yields quarterly changes fluctuation are explained by macro news. Tq na Tq np X X X X aniq = si,t wi pniq = si,t wi t=Tq−1 +1 i=1 t=Tq−1 +1 i=1 where Tq is the last day in quarter q and Tq X rxq = rxt = log(S&P 500Tq ) − log(S&P 500
Slide 17: Macro News and Quarterly Stock Returns
Macro News and Quarterly Stock Returns 17 / 37
Slide 18: Macro News and Quarterly Stock Returns
Macro News and Quarterly Stock Returns Our second set of results is based on the following regression: rxq = α + βaniq + γpniq + δmpsq + ϕrxq−1 + uq . aniq pniq mpsq rxq−1 R2 2 Radj 3.97 -2.05 -0.79 -0.06 34% 30% 3.36 -3.38 -0.76 -0.50 3.61 -0.86 -0.03 26% 23% 3.60 -0.81 -0.22 -1.38 -1.06 0.06 6% 2%
Slide 19: Signal vs. Noise
Signal vs. Noise The low daily R2 may be due to confounding mean-reverting factors that hide the macro “signal.” The persistency of a series yt can be gauged by considering var(yt − yt−h )/h, as a function of h (Cochrane, 1988). ▶ if var(yt − yt−h )/h is constant wrt. h then yt is a random walk. ▶ i
Slide 20: Signal vs. Noise
Signal vs. Noise Let’s define t X rxt−h,t = rxj j=t−h t X t X anit−h,t = anij pnit−h,t = pnij . j=t−h j=t−h For h from 1 to 66 we estimate: rxt−h,t = α(h) + β (h) anit−h,t + γ (h) pnit−h,t + ϵt−h,t 20 / 37
Slide 21: Signal vs. Noise
Signal vs. Noise ▶ Return is 1/h ∗ var(rxt−h,t ), ▶ Fit is 1/h ∗ var(c α(h) + βb(h) anit−h,t + γ rxt−h,t ) = 1/h ∗ var(b b(h) pnit−h,t ) ▶ Error is 1/h ∗ var(b ϵt−h,t ) Notice that 1/h var (rx b t−h,t ) R2 (h) := , b t−h,t ) + 1/h var (rxt−h,t − rx 1/h var (rx b t−h,t ) 21 / 37
Slide 22: State Dependency for the Activity Index
State Dependency for the Activity Index We our indexes with measure of business cycle or monetary policy stands xq : rxq = α + βaniq + β̃aniq ∗ xq + γpniq + γ̃pniq ∗ xq + δmpsq + θxq + ϕrxq−1 + uq Unemp. gap Output gap 1-year yield aniq 3.33 2.99 3.74 2.75 3.44 2.72 aniq ∗ xq 0.33 -0.49 0.38 0.85 -1
Slide 23: State Dependency for the Price Index
State Dependency for the Price Index rxq = α+βaniq +β̃aniq ∗xq +γpniq +γ̃pniq ∗xq +δmpsq +θxq +ϕrxq−1 +uq Unemp. gap Output gap 1-year yield pniq -2.13 -1.94 -1.34 -3.48 -3.09 -1.22 pniq ∗ xq 0.18 -0.23 -0.50 0.45 -0.62 -1.40 γ + γ̃ ∗ xq low -2.18 -1.31 -1.46 medium -2.05 -1.66 -1.86 high -1.62 -2.0
Slide 24: Firm's Reaction
Firm’s Reaction Following Ottonello and Winberry (2020): yi,q+h − yi,q−1 = αi,h + βh aniq + γh pniq + ϕh mpsq + Φ′1,h Zi,q−1 ... +Φ′2,h Wq−1 + δhF F iscalq + δhQ Qtrq + εi,q+h , Where: ▶ αi,h is a firm fixed effect. ▶ Zi,t−1 consists of leverage, log size, current assets, and yi,q − yi,q−1 . ▶ Wt−1
Slide 25: Firm's Reaction
Firm’s Reaction ▶ Revenues and Profitability + Sales, Income, Gross Margin ▶ Liquidity and Capital Structure + Cash, Receivables, Inventories, Book Equity, Liabilities ▶ Financing and Investment + Equity Issuance, Net Debt Issuance, Total Payout, R&D Expenditures, CAPX Yi,q+h − Yi,q−1 ∆Yi,q−1,q+h =
Slide 26: Revenues and Profitability
Revenues and Profitability (1) (2) (3) MKT Sales Income Gross Margin Panel A: Same quarter ani 7.531∗∗∗ 1.263∗∗∗ 0.240∗∗∗ 0.455∗∗∗ (3.228) (3.925) (2.751) (3.287) pni -4.362∗∗∗ 0.379 -0.025 -0.163 (-2.762) (1.119) (-0.627) (-1.619) Panel B: One quarter ahead ani 10.891∗∗∗ 2.677∗∗∗ 0.111 0.585∗ (2.92
Slide 27: Balance Sheet
Balance Sheet (1) (2) (3) (4) (5) Cash Rec. Invt. Book Equity Liabilities Panel A: Same quarter ani 1.215∗∗ 1.305∗∗∗ 0.179 1.152∗∗ 0.143 (2.644) (3.334) (1.106) (2.525) (0.667) pni -0.222 -0.062 0.460∗∗∗ 0.034 0.112 (-0.486) (-0.153) (2.807) (0.091) (0.616) Panel B: One quarter ahead ani 1.925∗∗∗ 2.
Slide 28: Financing and Investment
Financing and Investment (1) (2) (3) (4) (5) SSTK Net Debt Iss. Total P.O. R&D CAPX Panel A: Same quarter ani 0.113 0.022 0.006 -0.026 0.023∗∗∗ (0.910) (0.689) (0.237) (-1.495) (3.806) pni -0.166 -0.022 -0.007 0.024 0.008 (-1.278) (-0.616) (-0.379) (1.538) (1.222) Panel B: One quarter ahead ani 0.24
Slide 29: Macro Consequences of Firm-level Reactions
Macro Consequences of Firm-level Reactions Our second set of results is based IRFs from the following local projections, as in Ramey (2016): yq+h − yq−1 = αh + βh aniq + γh pniq + ϕh mpsq + Φ′1,h Zq−1 + εq+h , Where: ▶ αh is the constant. ▶ Zq−1 contains and yq − yq−1 , aniq−1 , pniq−1 and mpsq−1 ,
Slide 30: Hours Worked
Hours Worked 0.25 Activity Shock Price Shock horizons -0.1 -0.15 -0.2 5 10 15 horizons 30/37
Slide 31: Capacity Utilization
Capacity Utilization Activity Shock Price Schock horizons 10 horizons 15 31/37
Slide 32: Industrial Production
Industrial Production 32 / 37
Slide 33: Real Investment
Real Investment Activity Shock Price Schock horizons horizons 33 / 37
Slide 34: Unemployment Rate
Unemployment Rate Activity Shock fi 0.1 Price Schock horizons 10 horizons 15 34/37
Slide 35: Real GDP
Real GDP Activity Shock Price Schock 0.7 ‘ ‘ ‘ 0.6 0.5 0.4 0.3 0.2 0.1 5 10 15 5 10 15 horizons horizons 35 / 37
Slide 36: Conclusions
Conclusions We construct two indexes of macroeconomic news: an activity news index and a price news index. ▶ These indexes aggregate news for all macroeconomic releases for which surveys are available. ▶ The weights are based on a measure of market attention to macro releases. Independently on the s
Slide 37: Conclusions
Conclusions Through our indexes we show that the reaction of the equity market to macroeconomic news is rationalized by their propagation into the real economy. ▶ Firms reacts to macroeconomic surprises. A stream of favorable news is followed by mechanical firm-level responses (due to the better tha
Banque Nationale De Belgique — Technical Presentation (December 2024) on 6ix