El Salvador EMBI directed graphs
Monthly — is the current spread priced?
rich
El Salvador EMBI looks rich at 293 bp vs fair 336 bp (-43 bp tight).
The EMBI tightened 111 bp over the last 6 months (385 → 273 bp). Last month the print was roughly unchanged. Last month’s move was led by US high-yield spreads (+20 bp, wider); USD broad (-10 bp, tighter); the terms-of-trade gap (+10 bp, wider). It looks rich because the market print (293 bp) is tighter than macro fair value (336 bp, gap -43 bp).
- USD broad increased, which typically widens the EMBI spread. That does not appear to have been incorporated into the current spread.
| Current EMBI | 293 bp (2026-10-15) |
|---|---|
| Macro fair value | 336 bp (month-end 2026-09-30) |
| Rich / cheap | rich |
| Gap (current − fair) | -43 bp (-1.167σ) |
| Month-end residual | -63 bp |
| Residual σ | 37 bp |
| In-sample R² | 0.954 |
4 monthly print(s) may not yet have been tradable at the last EMBI date.
| Parent | Δ last month | Contrib bp | Typical print lag (d) |
|---|---|---|---|
| us_hy_oas (t) | 0.490 | +20.0 | 0 |
| dxy (t) | 2.401 | -10.4 | 0 |
| tot_gap (t) | -0.000 | +9.9 | 45 |
| tot (t) | 0.000 | +0.0 | 45 |
| debt_gap (t) | 0.000 | +0.0 | 35 |
| current_account (t) | 0.000 | +0.0 | 25 |
Has EMBI had time to trade the prints?
| Input | Est. release | Traded after print? | Days since |
|---|---|---|---|
| us_hy_oas | 2026-09-30 | yes | 15 |
| tot | 2026-11-14 | pending | -30 |
| tot_gap | 2026-11-14 | pending | -30 |
| dxy | 2026-09-30 | yes | 15 |
| debt_gap | 2026-11-04 | pending | -20 |
| current_account | 2026-10-25 | pending | -10 |
Monthly DAG
Narrative edges (strongest first)
| From | To | Lag | β (z) | Child R² | |
|---|---|---|---|---|---|
| public_debt_gdp | → | debt_gap | t | +1.000 | 0.116 |
| nominal_deficit | → | primary_balance | t | +0.816 | -0.100 |
| copper | → | tot_gap | t | -0.739 | -144.224 |
| copper | → | tot | t | -0.581 | -1296.742 |
| nominal_deficit | → | public_debt_gdp | t | -0.543 | -9.673 |
| debt_gap | → | embi_bra | t | +0.526 | -0.974 |
| public_debt_gdp | → | embi_bra | t | +0.526 | -0.974 |
| dxy | → | embi_bra | t | +0.518 | -0.974 |
| current_account | → | embi_bra | t | -0.506 | -0.974 |
| tot | → | current_account | t | +0.448 | -0.114 |
| tot_gap | → | embi_bra | t | -0.392 | -0.974 |
| tot | → | reserves | t | +0.382 | -7.491 |
| copper | → | exports_growth | t | +0.366 | -2.666 |
| tot_gap | → | current_account | t | +0.356 | -0.114 |
| primary_balance | → | public_debt_gdp | t | -0.355 | -9.673 |
| primary_balance | → | debt_gap | t | -0.355 | 0.116 |
| reserves | → | reserves_growth | t | +0.336 | 0.049 |
| copper | → | embi_bra | t | +0.313 | -0.974 |
| current_account | → | reserves_growth | t | +0.260 | 0.049 |
| tot | → | exports_growth | t | -0.214 | -2.666 |
| us_hy_oas | → | embi_bra | t | -0.206 | -0.974 |
| current_account | → | reserves | t | +0.176 | -7.491 |
| exports_growth | → | current_account | t | +0.110 | -0.114 |
| tot | → | primary_balance | t | -0.055 | -0.100 |
| reserves_growth | → | embi_bra | t | -0.054 | -0.974 |
| primary_balance | → | embi_bra | t | -0.052 | -0.974 |
| tot_gap | → | primary_balance | t | -0.044 | -0.100 |
| tot | → | embi_bra | t | -0.030 | -0.974 |
ΔEMBI engine (holdout score)
Ridge on ΔEMBI: 1 risk (HY or VIX or S&P) + 1 rates/USD + 1 commodity + 1 local + 1 stock. Ranked by holdout R², not quintile lift. Best: theory|lags=[0, 1, 2] · parents: us_hy_oas, dxy, copper, exports_growth, public_debt_gdp · 182 train / 20 test through 2026-10-31.
| From | To | Lag | β (z) | Child R² | |
|---|---|---|---|---|---|
| dxy | → | embi_bra | t | +0.171 | 0.356 |
| us_hy_oas | → | embi_bra | t | +0.149 | 0.356 |
| exports_growth | → | embi_bra | t | +0.109 | 0.356 |
| us_hy_oas | → | embi_bra | t−1 | +0.093 | 0.356 |
| copper | → | embi_bra | t−2 | +0.088 | 0.356 |
| exports_growth | → | embi_bra | t−1 | +0.077 | 0.356 |
| copper | → | embi_bra | t−1 | +0.059 | 0.356 |
| exports_growth | → | embi_bra | t−2 | -0.057 | 0.356 |
| dxy | → | embi_bra | t−2 | -0.047 | 0.356 |
| us_hy_oas | → | embi_bra | t−2 | +0.041 | 0.356 |
| copper | → | embi_bra | t | -0.032 | 0.356 |
| dxy | → | embi_bra | t−1 | -0.000 | 0.356 |
| public_debt_gdp | → | embi_bra | t | -0.000 | 0.356 |
| public_debt_gdp | → | embi_bra | t−1 | -0.000 | 0.356 |
| public_debt_gdp | → | embi_bra | t−2 | -0.000 | 0.356 |
| Method | Lags | OOS R² | σ-quintile | Lift vs naive | Naive middle | ±1 bin | 3-class | Sign | RMSE Δ | Edges | Parents |
|---|---|---|---|---|---|---|---|---|---|---|---|
| theory|lags=[0, 1, 2] ← best | [0, 1, 2] | 0.319 | 45.0% | -0.050 | 50.0% | 95.0% | 45.0% | 80.0% | 0.224 | 10 | 15 |
| pc|lags=[0, 1, 2] | [0, 1, 2] | 0.319 | 45.0% | -0.050 | 50.0% | 95.0% | 45.0% | 80.0% | 0.224 | 10 | 15 |
| pc_strict|lags=[0, 1, 2] | [0, 1, 2] | 0.319 | 45.0% | -0.050 | 50.0% | 95.0% | 45.0% | 80.0% | 0.224 | 10 | 15 |
| tan|lags=[0, 1, 2] | [0, 1, 2] | 0.319 | 45.0% | -0.050 | 50.0% | 95.0% | 45.0% | 80.0% | 0.224 | 9 | 15 |
| hybrid|lags=[0, 1, 2] | [0, 1, 2] | 0.319 | 45.0% | -0.050 | 50.0% | 95.0% | 45.0% | 80.0% | 0.224 | 16 | 15 |
| theory|lags=[0, 1] | [0, 1] | 0.216 | 50.0% | 0.000 | 50.0% | 90.0% | 50.0% | 75.0% | 0.241 | 10 | 10 |
| pc|lags=[0, 1] | [0, 1] | 0.216 | 50.0% | 0.000 | 50.0% | 90.0% | 50.0% | 75.0% | 0.241 | 10 | 10 |
| pc_strict|lags=[0, 1] | [0, 1] | 0.216 | 50.0% | 0.000 | 50.0% | 90.0% | 50.0% | 75.0% | 0.241 | 10 | 10 |
| tan|lags=[0, 1] | [0, 1] | 0.216 | 50.0% | 0.000 | 50.0% | 90.0% | 50.0% | 75.0% | 0.241 | 9 | 10 |
| hybrid|lags=[0, 1] | [0, 1] | 0.216 | 50.0% | 0.000 | 50.0% | 90.0% | 50.0% | 75.0% | 0.241 | 16 | 10 |
| theory|lags=[1] | [1] | -0.076 | 35.0% | -0.150 | 50.0% | 90.0% | 35.0% | 70.0% | 0.282 | 10 | 5 |
| pc|lags=[1] | [1] | -0.076 | 35.0% | -0.150 | 50.0% | 90.0% | 35.0% | 70.0% | 0.282 | 10 | 5 |
| pc_strict|lags=[1] | [1] | -0.076 | 35.0% | -0.150 | 50.0% | 90.0% | 35.0% | 70.0% | 0.282 | 10 | 5 |
| tan|lags=[1] | [1] | -0.076 | 35.0% | -0.150 | 50.0% | 90.0% | 35.0% | 70.0% | 0.282 | 9 | 5 |