▪️ Attached Summary of differences
▪️ Since 2000, overall CPI ~11% higher than PCE
▪️ Definition:
🔹 CPI: Out-of-pocket spending by non-institutional Urban Consumers
🔹 PCE: Includes Rural & all personal sector
1/12
Four Sources of differences: Scope, Formula, Weight, Others Effects
1⃣ Scope Effect:
🔹 CPI: Consumer Price Index => Survey of Households
🔹 PCE: Personal Consumption Expenditure => Survey of Businesses
▪️ 25% of PCE spending not captured by CPI
2/12
▪️ PCE includes spending by Govt, Firms, Non-Profits on behalf of Households
- E.G: Medical spending = Direct purchases by Consumers + Spending on medical goods & services by Medicare OR Employer's Health Insurance
- E.G: Public school education not an out-of-pocket spending
3/12
- E.G: Imputed cost of fin servs that do not involve out-of-pocket spending => not in CPI
2⃣ Formula Effect:
- PCE's Fisher Ideal Formula better reflects consumer behaviour of substituting away from products with rising prices
- CPI's Laspeyres holds weights fixed for 2yrs
4/12
3⃣ Weight Effect:
- PCE's broader scope implies smaller weight to items common to both
- Housing ~33% weight in CPI but ~16% in PCE
- Gasoline/Energy heavier in CPI than PCE
- Rent/Gasoline Inflation would tend to lift CPI above PCE inflation
4⃣ Other Effects: Seasonal Adjustments & Revisions
🔹 CPI: Methodological improvements not applied to historical data => in a way good as CPI is used for Indexing, Social Security benefits, financial instruments
🔹 PCE: Improvements/Revisions applied to entire history
6/12
▪️ Weight & Scope Effect dominant drivers followed by Formula Effect
▪️ Weight Effect (+ve) has tended to push CPI above PCE
▪️ Q4'08-Q1'09 saw collapse in Energy prices => pushed down CPI more than PCE given Energy's heavier weight in CPI; similar in 2014-15 & dip in 2020
7/12
Four Items leading to key differences:
🔹 Healthcare=>PCE has third party expenses
🔹 Housing=>methodology & definition differences
🔹 Education=>PCE has third party expenses
🔹 Vices (Alcohol, etc)=>Consumers (CPI) may not admit, while Businesses not shy of reporting (PCE)
8/12
▪️ Median Difference b/w YoY Inflation Rates (CPI v/s PCE) has come off recently since COVID
▪️ But in April, CPI again spiked above PCE given massive jump in CPI-heavy Energy prices
Median Difference:
21yr history: 0.30%
5yr history: 0.30%
Post-COVID: 0.05%
9/12
So why Fed likes PCE over CPI?
Three key reasons summarized in report to Congress in 2000 when switch was made
1⃣ PCE's better substitution effect v/s fixed weight CPI (Ref 4/12 above)
2⃣ PCE's broader coverage (Ref 2-3/12)
3⃣ Ability to revise historical PCE (Ref 6/12)
10/12
Fed actually like "Core" PCE even more (excludes Food/Energy) as Headline more volatile => another debate (may require another thread) whether Fed should focus on Headline or Core
Since 2000, Headline CPI v/s Core PCE Median difference = 0.40% (handy conversion guide)
11/12
Wow, look at the volatility of Energy Inflation => have to go in favour of Core
▪️ #USDINR last 74.35 => ~1.4% off 75.35 highs => RBI's persistent #USD selling above 75.00 => with soft DXY, s/t consolidation in 74.00-75.35?
▪️ Risk Reversals, good gauge of nervousness, off highs (+1.6=>+0.9 vol) => less demand for USD Calls
1/11
▪️ Various economists revised India's GDP forecast lower
▪️ Good summary by @latha_venkatesh below
▪️ RBI GDP Projection +10.5% yoy FY 21/22 (Apr MPC)
▪️ Chart below: graphical overview of GDP trajectory - not that bad but whether worse yet to come?
▪️ When GDP collapses =>Trade Deficit tends to improve=>lower imports on poor aggregate demand
▪️ Q2 Apr-Jun'20=>massive reduction in trade deficit as GDP collapsed
▪️ Assuming only mild GDP hit in this COVID wave, associated trade deficit improvement should also be smaller
▪️ That +1.0% gap would be largest Headline v/s Core CPI gap since 2011 on huge Energy driven base effects; 5yr avrg gap -0.20%, 2yr avrg -0.37%
1/4
▪️ Chart: Energy Inflation indeed so volatile
▪️ Food Inflation key contributor since COVID - likely to ease off marginally on base effects
▪️ Energy base effect => massive spike to Headline over Apr/May => Headline vs Core likely to diverge further
▪️ Energy Inflation, negative recently, was +2.4% Feb, could jump to as high as +8.5% yoy in Mar & +20% in April
▪️ Energy Index collapsed Apr/May'20 but Food Index jumped => opposite base effects
US Swaption/Rates Vol off highs but still elevated
Reasons:
▪️ Economic data ahead; CPI tom; Fed data dependent
▪️ UST supply, 3y 10y 30y auctions this wk
▪️ US ylds, off highs, still close to higher end => risk premium in vols
▪️ Market v/s Fed disconnect on Rate Hike
1/4
▪️ SEP-based June FOMC important => till then would have got two more NFPs & Inflation data till May incorporating Apr/May base effect + stimulus based spikes => little kink around 3m in vol curve
▪️ Vol Skew mildly softer but still topside nervousness (convexity hedging)
2/4
▪️ Chart: 3m Expiry Vol across Swap Tenors => Front end swap tenors least volatile as anchored by Fed Fund/OIS/T-Bill yields
▪️ Chart: 3m10y Swaption Vol v/s 10yr US Yield: higher yield => higher vol as reflected in Vol Skew
▪️ Named 'Government Securities Acquisition Programme', G-SAP 1.0
▪️ Commits to specific amt of G-Sec purchases (INR 1.0 trn in Q1 FY2021-22)
▪️ At annualized INR 4 trn, that is ~7% of RBI Balance Sheet BS (Fed's $1.44 trn QE => ~19% BS)
1/5
G-SAP = YCC?
▪️ While RBI prefers 10yr yield to be ard 6.0%, formal Yield Curve Control YCC targets specific yld for specific tenor (BOJ Japan & RBA Australia) => unknown amt of bond purchase to meet yld target
▪️ RBI's G-SAP more QE (pre-announced amt) than YCC
G-SAP v/s OMO?
▪️ Practically not much diff but RBI generally does not commit to purchase amt under OMO
▪️ OMO more for liquidity mgmt => inject as well as absorb liquidity
▪️ QE purchase necessarily injects liquidity; targets backend ylds => BS expansion => Reserve Money ⬆️
#2 Nominal Interest Rates (x-axis) => 10yr US Treasury Yield
1/10
Q=Quadrant
Q1: Risk-On-Lower Real Rates => Short USD
Q1: Risk-On-Higher Real Rates => Long USD
Q2: Goldilocks => Risk-On-Lower Nominal/Real => Short USD
Q3: Risk Off => Lower Rates => Equity Sell off/Bond Rally => Long USD
Q4: Risk Off => Equity & Bond Sell off => Long USD
2/10
Performance since COVID, Mar'20
Q1:
- Higher BE/Risk On + Higher Nominals => 31% of all observations
- Good success (85%) on Short USD under lower Real (18.4% of obs)
- But higher Real under Risk On only 20% success on Long USD
- So despite higher Real, USD lower on Risk On
3/10
BE = inflation level that makes investor indifferent b/w buying UST or TIPS
= (approx) market-based measure of risk neutral inflation expectations
= TIPS indexed to non-seasonally adjusted #CPI