Skip to contents

Margin of conservatism, by category

Source: R/pd.R
scr_moc.Rd

Appends entries to the MoC ledger of an scr_grades() object. Category "C" (general estimation error) is quantified: "ci_timeseries" takes the upper bound of a one-sided level interval of the long-run average from the cohort series, \(t_{q, T-1}\, sd(DR_t)/\sqrt{T}\) per grade; "ci_binomial" uses \(z_q \sqrt{PD(1-PD)/n}\) on the obligors (or obligor-years when a series exists); "bootstrap" resamples the outcomes of the sample within each grade (drawn as the resampled default rate, Binomial(n, DR) / n, its exact distribution) and takes the level quantile of the default rate above the estimate. Categories "A" (data and methodological deficiencies) and "B" (changes in standards or environment) are expert quantities: value (one number or one per grade, in PD units) and a non-empty reason are mandatory. The ledger is append-only: A and B entries accumulate, a new C supersedes the previous one (kept with active = FALSE).

Usage

scr_moc(
  x,
  category = c("A", "B", "C"),
  method = NULL,
  level = NULL,
  value = NULL,
  reason = NULL,
  dr = NULL,
  n_boot = 200L,
  seed = NULL
)

Arguments

x

An scr_grades() object.

category

"A", "B" or "C".

method

For "C": "ci_timeseries", "ci_binomial" or "bootstrap"; NULL reads config$pd_moc_method.

level

One-sided confidence level; NULL reads config$pd_moc_level.

value

For "A"/"B": the add-on in PD units, length 1 or one per grade.

reason

Justification (mandatory for "A"/"B").

dr

Optional scr_dr by grade for "ci_timeseries", keyed by the final grades of x; NULL uses the series already stored in x$dr by scr_grades().

n_boot, seed

Bootstrap resamples and seed for "bootstrap".

Value

The scr_grades object with the entries appended to moc.

See also

Other irb-pd: predict.scr_grades(), predict.scr_pd(), scr_calibrate(), scr_grades(), scr_master_scale(), scr_migration(), scr_pd(), scr_pd_pit_ttc(), scr_pd_validate()

Examples

cfg <- scr_config(verbose = FALSE, nthread = 1, use_ranger = FALSE,
                  use_lightgbm = FALSE, xgb_rounds = 40, n_boot = 10)
res <- scr_select(scr_demo, "default", config = cfg, drop = c("id", "churn"),
                  date_col = "ref_date")
sc <- scr_scorecard(res)
gr <- scr_grades(sc, n_grades = 6, min_defaults = 10)
gr <- scr_moc(gr, "C", method = "ci_binomial")
gr <- scr_moc(gr, "A", value = 0.002, reason = "missing unlikeliness-to-pay trigger before 2024")
gr$moc
#>       id category      method level grade      pd_be      value
#>    <int>   <char>      <char> <num> <int>      <num>      <num>
#> 1:     1        C ci_binomial  0.95     1 0.03382664 0.01367268
#> 2:     1        C ci_binomial  0.95     2 0.12800000 0.02457568
#> 3:     1        C ci_binomial  0.95     3 0.26567164 0.03969379
#> 4:     1        C ci_binomial  0.95     4 0.36956522 0.08277496
#> 5:     2        A      manual    NA     1 0.03382664 0.00200000
#> 6:     2        A      manual    NA     2 0.12800000 0.00200000
#> 7:     2        A      manual    NA     3 0.26567164 0.00200000
#> 8:     2        A      manual    NA     4 0.36956522 0.00200000
#>                                             reason active       date
#>                                             <char> <lgcl>     <char>
#> 1:  estimation error, ci_binomial at 95% one-sided   TRUE 2026-10-02
#> 2:  estimation error, ci_binomial at 95% one-sided   TRUE 2026-10-02
#> 3:  estimation error, ci_binomial at 95% one-sided   TRUE 2026-10-02
#> 4:  estimation error, ci_binomial at 95% one-sided   TRUE 2026-10-02
#> 5: missing unlikeliness-to-pay trigger before 2024   TRUE 2026-10-02
#> 6: missing unlikeliness-to-pay trigger before 2024   TRUE 2026-10-02
#> 7: missing unlikeliness-to-pay trigger before 2024   TRUE 2026-10-02
#> 8: missing unlikeliness-to-pay trigger before 2024   TRUE 2026-10-02