IESET.
Hypotheses·monetary·forward_guidance_term_structure_effect

Post-2008 mainstream-central-bank forward guidance affected the term structure of interest rates beyond what pure-signal rational-expectations models would predict, consistent with NK sticky-information models.

SUPPORTEDengine/runs/forward_guidance_term_structure_effect

SUPPORTED (US-FOMC daily channel only) — mean |ΔDGS2| on 5 FG days = 6.6 bp; placebo mean |Δ| = 1.3 bp (ratio 4.95x); DGS5 ratio = 4.06x. Mean |ΔDGS2| clears the 5bp magnitude bar AND the 2x placebo bar; DGS5 also clears the 1.5x informative bar. ECB/BoE/BoJ legs and intraday discrimination of sticky-info vs RE-signal remain METHOD_VALID data gaps.

confidence cueThis is a clear pass for the claim as written. It still applies only to this sample, period, and method.

policy briefNeeds review

In ordinary language

In plain terms, this asks whether forward guidance announcement event is actually linked to better or worse yield 2y from 2008 to 2017.

plain answer

The data clearly moved in the predicted direction. mean |ΔDGS2| on 5 FG days = 6.6 bp; placebo mean |Δ| = 1.3 bp (ratio 4.95x); DGS5 ratio = 4.06x.

why it matters

This matters because monetary claims should change belief only when they survive a pre-declared empirical test.

how the test works

It compares 4 country or place units from 2008 to 2017, using a event study design.

what was measured
What changed
  • Forward guidance announcement event
  • Ois implied policy path change
What we checked
  • Yield 2y
  • Yield 5y
  • Yield 10y
what this does not prove

A single test is not the whole truth. It narrows the claim under a specific sample, time period, and method. Strong policy conclusions need the pattern to survive nearby tests, alternative data, and serious objections.

verification

No evidence packet has been generated yet.

Results

engine/runs/forward_guidance_term_structure_effect
1007550250200820132017USAGBRDEUJPN
illustrative sketch · run pending
No coefficients yet. When the model fires, this chart will show yield_2y across 4 sampled countries over 2008–2017.
The shapes above are stylised — none of the lines are real data.
Placeholder for forward_guidance_term_structure_effect. Published chart will be generated from engine/runs/forward_guidance_term_structure_effect/chart_data.json.

Who has skin in the game — schools predicting on this

1 school list this hypothesis as a test of their position. The chips below are school-level scoreboard outcomes, not a second hypothesis verdict.

hypothesis verdict vs scoreboard outcome

The banner verdict judges this hypothesis as written. The scoreboard asks whether each school's polarity-corrected prediction was right. Raw status is not a school win: SUPPORTED supports schools that needed SUPPORTED, but refutes schools that needed REFUTED.

New Keynesian (mainstream synthesis)wins if SUPPORTEDschool supported

Pre-registration

pre-registered
first-spec commit bae09ab · 2026-04-29T22:09:42Z

Post-2008 mainstream-central-bank forward guidance affected the term structure of interest rates beyond what pure-signal rational-expectations models would predict, consistent with NK sticky-information models.

Falsification criterion — what would disprove this

set before the run · honoured after

This hypothesis is considered falsified if:

The hypothesis is falsified if the preregistered empirical test shows the opposite direction of the claim at conventional significance (p < 0.05), or if the primary outcome measure does not move in the claimed direction.

formal test & threshold
test:      forward_guidance_term_structure_effect_placeholder_test

Method

Template
event_study
Clustering
announcement
Sample
4 countries · 2008 – 2017
Evidence type
associational

Stub-level estimator pin for runnability audit. High-frequency event-study around major Fed/ECB/BoE/BoJ forward-guidance announcements 2008-2017; outcome is the change in 2y/5y/10y nominal yields and OIS-implied policy paths in 30-minute windows. Compares observed term-structure response against pure-signal RE benchmarks. Falsification rule and variables block remain to be filled when this stub is promoted from draft.

Data

VariableSourceTransform
yield_2y
outcome
fred:DGS2tier 1
boe:IUDBEDRtier 1
ecb:financial_markets_yields_2yrtier 1
boj:policy_ratetier 1
level_change_bps
yield_5y
outcome
fred:DGS5tier 1
ecb:financial_markets_yields_5yrtier 1
level_change_bps
yield_10y
outcome
fred:DGS10tier 1
ecb:financial_markets_yields_10yrtier 1
level_change_bps
forward_guidance_announcement_event
treatment
constructed:hand-coded event list — Fed FOMC (2008-12, 2011-08, 2012-08, 2013-12, 2015-09), ECB GC (2013-07, 2014-09), BoE MPC (2013tier 5
event_date
ois_implied_policy_path_change
treatment
fred:OIStier 1
ecb:financial_markets_money_market_OIStier 1
level_change_bps
vix
control
fred:VIXCLStier 1
level
surprise_macro_release_indicator
control
constructed:binary = 1 if FOMC/MPC/GC day coincided with NFP/CPI releasetier 5
binary

● ready  · ● pending  · ● reconstruct-needed

Detailed result card

Forward-guidance term-structure event study (post-2008 FOMC)

Verdict: SUPPORTED (US-FOMC daily channel only) — mean |ΔDGS2| on 5 FG days = 6.6 bp; placebo mean |Δ| = 1.3 bp (ratio 4.95x); DGS5 ratio = 4.06x. Mean |ΔDGS2| clears the 5bp magnitude bar AND the 2x placebo bar; DGS5 also clears the 1.5x informative bar. ECB/BoE/BoJ legs and intraday discrimination of sticky-info vs RE-signal remain METHOD_VALID data gaps.

Summary

  • Hand-coded FOMC forward-guidance event dates (n=5): 2008-12-16, 2011-08-09, 2012-09-13, 2013-12-18, 2015-09-17.
  • Mean absolute 1-day Δ DGS2 on FG days: 6.6 bp (n=5).
  • Mean absolute 1-day Δ DGS2 on calendar-matched non-FOMC Wednesdays: 1.3 bp (n=21).
  • FG/placebo ratio (DGS2): 4.95x (threshold 2.0x).
  • DGS5 mean |Δ|: FG 11.2 bp vs placebo 2.8 bp (ratio 4.06x; informative threshold 1.5x).
  • DGS10 mean |Δ|: FG 10.2 bp.

Per-event 1-day yield changes (pp)

| Event | DGS2 Δ | DGS5 Δ | DGS10 Δ | |---|---|---|---| | 2008-12-16 | -0.100 | -0.160 | -0.160 | | 2011-08-09 | -0.080 | -0.200 | -0.200 | | 2012-09-13 | -0.010 | -0.050 | -0.020 | | 2013-12-18 | -0.020 | +0.030 | +0.040 | | 2015-09-17 | -0.120 | -0.120 | -0.090 |

Method

Daily-frequency US-FOMC forward-guidance event study. For each of the 5 hand-coded FG announcement dates, compute the 1-day change in DGS2/DGS5/DGS10 (post-event-day close minus prior-trading-day close, in percentage points). Compare the mean absolute FG-day change against a calendar-matched placebo of all non-FOMC Wednesdays in the same months. The dispositive primary is mean |ΔDGS2|.

METHOD_VALID data gaps

  • Intraday data: the original spec asked for 30-minute windows. Only daily FRED is on disk. Daily windows are noisier and cannot attribute movement specifically to the FG headline.
  • Non-US central banks: ECB / BoE / BoJ yield series and FG event lists are not in the current vintage tree.
  • Pure-signal RE benchmark: a calibrated NK model is required to discriminate sticky-information from rational-expectations channels per the original spec; no such model in repo. The daily test only detects whether the term structure moved materially on FG days.

Data

  • fred:DGS2 — primary outcome (US 2y treasury yield, daily)
  • fred:DGS5 — informative outcome (US 5y, daily)
  • fred:DGS10 — informative outcome (US 10y, daily)
  • fred:VIXCLS — backdrop control (CBOE VIX, daily)

Notes

Origin is auto-generated coverage-gap stub seeded from New Keynesian sticky-information framing of post-2008 forward-guidance term-structure effects. Human review required before promotion.

Authored framework. Read the transparency note.