MaxGhenis · Oct 1, 2026 · Oct 4, 2026 · Oct 12, 2026 · Oct 12, 2026
diff --git a/changelog.d/was-wealth-components.added.md b/changelog.d/was-wealth-components.added.md
@@ -0,0 +1 @@
+Export `private_pension_wealth`, `directly_held_shares`, `unit_and_investment_trusts`, `stocks_and_shares_isa` and `cash_isa` from the Wealth and Assets Survey, and impute debt secured on non-main property and land (`other_residential_property_secured_debt`, `non_residential_property_secured_debt`, `owned_land_secured_debt`).
diff --git a/changelog.d/was-wealth-split.fixed.md b/changelog.d/was-wealth-split.fixed.md
@@ -0,0 +1 @@
+Split private pension wealth out of `corporate_wealth`, which now holds only shares, unit and investment trusts and stocks and shares ISAs, and add buy-to-let property to `other_residential_property_value`.
diff --git a/docs/imputations.md b/docs/imputations.md
@@ -19,9 +19,9 @@ The imputations are applied in this order (dependencies noted):
 
 ## Wealth Imputation
 
-**Source:** Wealth and Assets Survey (WAS) Round 7 (2018-2020)
+**Source:** Wealth and Assets Survey (WAS) Round 8 (April 2020 to March 2022), household file
 
-Imputes household wealth components using demographic and income predictors.
+Imputes household wealth components using demographic and income predictors. The targets are imputed as a chain: each is fitted on the predictors and every earlier target.
 
 ### Predictors
 | Variable | Description |
@@ -39,18 +39,31 @@ Imputes household wealth components using demographic and income predictors.
 | `region` | UK region |
 
 ### Outputs
-| Variable | Description |
-|----------|-------------|
-| `owned_land` | Value of owned land |
-| `property_wealth` | Total property wealth |
-| `corporate_wealth` | Shares, pensions, investment ISAs |
-| `gross_financial_wealth` | Total financial assets |
-| `net_financial_wealth` | Financial assets minus liabilities |
-| `main_residence_value` | Value of main home |
-| `other_residential_property_value` | Value of other properties |
-| `non_residential_property_value` | Value of non-residential property |
-| `savings` | Savings account balances |
-| `num_vehicles` | Number of vehicles owned |
+| Variable | WAS round 8 source | Description |
+|----------|--------------------|-------------|
+| `owned_land` | `DVLUKValR8_sum` | Value of land in the UK |
+| `property_wealth` | `DVPropertyR8` | Gross value of all property, main residence included |
+| `private_pension_wealth` | `totalpenr8_aggr` − `dvvaldbt_scaper8_aggr` | Private pension wealth other than current-employment defined benefit rights: defined contribution pots, retained rights, pensions in payment |
+| `directly_held_shares` | `DVFShUKVR8_aggr` + `DVFESHARESR8_aggr` | UK shares and employee shares and options, held outside ISAs and pooled funds |
+| `unit_and_investment_trusts` | `DVFCollVR8_aggr` | Unit trusts and investment trusts (one survey question) |
+| `stocks_and_shares_isa` | `DVIISAVR8_aggr` | Stocks and shares (investment) ISAs |
+| `corporate_wealth` | derived | `directly_held_shares` + `unit_and_investment_trusts` + `stocks_and_shares_isa`, summed after imputation |
+| `gross_financial_wealth` | `HFINWR8_SUM` | Total financial assets |
+| `net_financial_wealth` | `HFINWNTR8_Sum` | Financial assets minus liabilities |
+| `main_residence_value` | `DVHValueR8` | Value of main home |
+| `other_residential_property_value` | `DVHseValR8_sum` + `DVBltValR8_sum` | Second homes and buy-to-let property |
+| `non_residential_property_value` | `DVBlDValR8_sum` | Other buildings, such as shops and garages |
+| `savings` | `DVSaValR8_aggr` | Savings account balances (cash ISAs excluded) |
+| `num_vehicles` | `vcarnr8` | Number of vehicles owned |
+| `student_loan_balance` | `Tot_LosR8_aggr` − `Tot_los_exc_SLCR8_aggr` | Student Loans Company balance, allocated to people |
+| `cash_isa` | `DVCISAVR8_aggr` | Cash ISAs |
+| `other_residential_property_secured_debt` | `DVHseDebtR8_sum` + `DVBLtDebtR8_sum` | Mortgages and loans secured on second homes and buy-to-let property |
+| `non_residential_property_secured_debt` | `DVBldDebtR8_sum` | Mortgages and loans secured on other buildings |
+| `owned_land_secured_debt` | `DVLUKDebtR8_sum` | Loans secured on UK land |
+
+`corporate_wealth` is the exact sum of its three components on every household; private pension wealth is not part of it, because pension rights are disregarded capital in every means test. A secured debt is zero wherever the household holds none of the asset it is secured on. Land and property overseas (`DVLOSValR8_sum`), other real estate (`DVOPrValR8_sum`) and overseas shares (`DVFShOSVR8_aggr`) are not exported separately; `property_wealth` and the financial wealth totals include them.
+
+Each target draws its quantile with its own seed. microimpute otherwise uses one seed for every target, so a household drawn high for one asset is drawn high for every later target conditioned on it, which makes sparse targets such as the debt secured on land near-certain for the households that hold the asset.
 
 ---
 

diff --git a/policyengine_uk_data/datasets/imputations/wealth.py b/policyengine_uk_data/datasets/imputations/wealth.py
@@ -45,10 +45,17 @@
     "region",
 ]
 
+# Chain order matters: each target is fitted on the predictors plus every
+# earlier target. Private pension wealth takes the slot the old pension-laden
+# corporate_wealth held, the share-like components follow it, and each
+# secured debt comes after the asset it is secured on.
 IMPUTE_VARIABLES = [
     "owned_land",
     "property_wealth",
-    "corporate_wealth",
+    "private_pension_wealth",
+    "directly_held_shares",
+    "unit_and_investment_trusts",
+    "stocks_and_shares_isa",
     "gross_financial_wealth",
     "net_financial_wealth",
     "main_residence_value",
@@ -57,17 +64,67 @@
     "savings",
     "num_vehicles",
     "student_loan_balance",
+    "cash_isa",
+    "other_residential_property_secured_debt",
+    "non_residential_property_secured_debt",
+    "owned_land_secured_debt",
 ]
 
+# corporate_wealth is not imputed: it is the sum of its imputed components,
+# so the identity holds on every household.
+CORPORATE_WEALTH_COMPONENTS = (
+    "directly_held_shares",
+    "unit_and_investment_trusts",
+    "stocks_and_shares_isa",
+)
+
+# Debt secured on each capital asset (UC Regs 2013 reg. 49(1)(b) and its
+# legacy equivalents deduct it from that asset's value only).
+SECURED_DEBT_ASSETS = {
+    "other_residential_property_secured_debt": "other_residential_property_value",
+    "non_residential_property_secured_debt": "non_residential_property_value",
+    "owned_land_secured_debt": "owned_land",
+}
+
+# microimpute draws each target's quantile from the same seed, so a household
+# gets the same quantile for every target. Chained on earlier draws, that makes
+# sparse targets near-certain for households drawn high for a related asset
+# (nearly every imputed land holder had debt secured on the land). Each target
+# gets its own seed so the draws are independent.
+QUANTILE_DRAW_SEEDS = {
+    variable: 1_000 + index for index, variable in enumerate(IMPUTE_VARIABLES)
+}
+
+# WAS round 8 sources of the targets built from more than one column. Part of
+# the model metadata, so a change of definition retrains a cached model.
+DERIVED_TARGETS = {
+    "private_pension_wealth": ("totalpenr8_aggr", "-dvvaldbt_scaper8_aggr"),
+    "directly_held_shares": ("DVFShUKVR8_aggr", "DVFESHARESR8_aggr"),
+    "other_residential_property_value": ("DVHseValR8_sum", "DVBltValR8_sum"),
+    "other_residential_property_secured_debt": (
+        "DVHseDebtR8_sum",
+        "DVBLtDebtR8_sum",
+    ),
+    "student_loan_balance": ("Tot_LosR8_aggr", "-Tot_los_exc_SLCR8_aggr"),
+}
+
 WAS_RENAMES = {
     "R8xshhwgt": "household_weight",
     # Components for estimating land holdings.
     "DVLUKValR8_sum": "owned_land",  # In the UK.
+    "DVLUKDebtR8_sum": "owned_land_secured_debt",
     "DVPropertyR8": "property_wealth",
+    # UK shares (listed or not) and employee shares and options, held outside
+    # ISAs and pooled funds.
     "DVFESHARESR8_aggr": "emp_shares_options",
     "DVFShUKVR8_aggr": "uk_shares",
-    "DVIISAVR8_aggr": "investment_isas",
-    "DVFCollVR8_aggr": "unit_investment_trusts",
+    # Investment ISAs: the survey's stocks and shares ISA question.
+    "DVIISAVR8_aggr": "stocks_and_shares_isa",
+    "DVCISAVR8_aggr": "cash_isa",
+    # Unit trusts and investment trusts: one survey question.
+    "DVFCollVR8_aggr": "unit_and_investment_trusts",
+    # Total private pension wealth and its current-employment defined benefit
+    # part, both valued on the SCAPE basis.
     "totalpenr8_aggr": "pensions",
     "dvvaldbt_scaper8_aggr": "db_pensions",
     # Predictors for fusing to FRS.
@@ -88,12 +145,17 @@
     # Other columns for reference.
     "DVLOSValR8_sum": "non_uk_land",
     "HFINWNTR8_Sum": "net_financial_wealth",
-    "DVLUKDebtR8_sum": "uk_land_debt",
     "HFINWR8_SUM": "gross_financial_wealth",
     "TotalWlthR8": "wealth",
     "DVhvalueR8": "main_residence_value",
-    "DVHseValR8_sum": "other_residential_property_value",
+    # Gross values of property other than the main residence, and the
+    # mortgages and loans secured on each class.
+    "DVHseValR8_sum": "second_homes_value",
+    "DVHseDebtR8_sum": "second_homes_debt",
+    "DVBltValR8_sum": "buy_to_let_value",
+    "DVBLtDebtR8_sum": "buy_to_let_debt",
     "DVBlDValR8_sum": "non_residential_property_value",
+    "DVBldDebtR8_sum": "non_residential_property_secured_debt",
     "DVTotinc_bhcR8": "household_net_income",
     "DVSaValR8_aggr": "savings",
     "vcarnr8": "num_vehicles",
@@ -142,16 +204,18 @@ def generate_was_table(was: pd.DataFrame):
 
     was["is_renting"] = was["is_renter"] == 1
 
-    was["non_db_pensions"] = was.pensions - was.db_pensions
-    was["corporate_wealth"] = was[
-        [
-            "non_db_pensions",
-            "emp_shares_options",
-            "uk_shares",
-            "investment_isas",
-            "unit_investment_trusts",
-        ]
-    ].sum(axis=1)
+    # Private pension wealth other than current-employment defined benefit
+    # rights: pension rights are disregarded capital in every means test, so it
+    # is kept out of corporate_wealth.
+    was["private_pension_wealth"] = was.pensions - was.db_pensions
+    was["directly_held_shares"] = was.uk_shares + was.emp_shares_options
+    was["corporate_wealth"] = was[list(CORPORATE_WEALTH_COMPONENTS)].sum(axis=1)
+    was["other_residential_property_value"] = (
+        was.second_homes_value + was.buy_to_let_value
+    )
+    was["other_residential_property_secured_debt"] = (
+        was.second_homes_debt + was.buy_to_let_debt
+    )
     was["student_loan_balance"] = was["total_loans"] - was["total_loans_exc_slc"]
     was["region"] = was["region"].map(REGIONS)
     return was
@@ -162,6 +226,8 @@ def generate_was_table(was: pd.DataFrame):
     "was_household_tab_filename": CURRENT_WAS_RELEASE.household_tab_filename,
     "predictor_variables": tuple(PREDICTOR_VARIABLES),
     "impute_variables": tuple(IMPUTE_VARIABLES),
+    "derived_targets": tuple(DERIVED_TARGETS.items()),
+    "quantile_draw_seeds": tuple(QUANTILE_DRAW_SEEDS.items()),
 }
 
 
@@ -269,6 +335,32 @@ def _allocate_student_loan_balance_to_people(
     return balances
 
 
+def derive_wealth_outputs(output_df: pd.DataFrame) -> pd.DataFrame:
+    """
+    Add the outputs derived from imputed targets.
+
+    The share-like components are clipped at zero and summed into
+    corporate_wealth, so the identity holds on every household. A secured debt
+    is kept only where the household holds the asset it is secured on.
+    """
+    output_df = output_df.copy()
+    for column in (*CORPORATE_WEALTH_COMPONENTS, *SECURED_DEBT_ASSETS):
+        output_df[column] = output_df[column].clip(lower=0)
+    output_df["corporate_wealth"] = output_df[list(CORPORATE_WEALTH_COMPONENTS)].sum(
+        axis=1
+    )
+    for debt, asset in SECURED_DEBT_ASSETS.items():
+        output_df[debt] = output_df[debt].where(output_df[asset] > 0, 0.0)
+    return output_df
+
+
+def use_independent_quantile_draws(model: QRF) -> QRF:
+    """Give each imputed target its own quantile-draw seed."""
+    for variable, seed in QUANTILE_DRAW_SEEDS.items():
+        model.model.models[variable].seed = seed
+    return model
+
+
 def save_imputation_models():
     """
     Train and save wealth imputation model.
@@ -290,6 +382,7 @@ def save_imputation_models():
         was[PREDICTOR_VARIABLES],
         was[IMPUTE_VARIABLES],
     )
+    use_independent_quantile_draws(wealth)
     wealth.save(get_wealth_model_path())
     return wealth
 
@@ -338,7 +431,7 @@ def impute_wealth(dataset: UKSingleYearDataset) -> UKSingleYearDataset:
     input_df["region"] = input_df["region"].replace(
         "NORTHERN_IRELAND", "WALES"
     )  # WAS doesn't sample NI -> put NI households in Wales (closest aggregate)
-    output_df = model.predict(input_df)
+    output_df = derive_wealth_outputs(model.predict(input_df))
 
     for column in output_df.columns:
         if column == "student_loan_balance":

diff --git a/policyengine_uk_data/storage/uprating_factors.csv b/policyengine_uk_data/storage/uprating_factors.csv
@@ -6,18 +6,23 @@ bsp_reported,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1
 capital_gains,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 capital_gains_before_response,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 carers_allowance_reported,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
+cash_isa,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 child_benefit_reported,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 child_tax_credit_reported,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 childcare_expenses,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 clothing_and_footwear_consumption,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 communication_consumption,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 corporate_wealth,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 diesel_spending,1.0,1.138,1.57,1.539,1.589,1.54,1.507,1.46,1.415,1.355,1.284,1.284,1.284,1.284,1.284
+directly_held_shares,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 dividend_income,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 domestic_energy_consumption,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 education_consumption,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 employee_pension_contributions,1.0,1.059,1.127,1.205,1.261,1.308,1.337,1.365,1.396,1.431,1.431,1.431,1.431,1.431,1.431
 employer_pension_contributions,1.0,1.059,1.127,1.205,1.261,1.308,1.337,1.365,1.396,1.431,1.431,1.431,1.431,1.431,1.431
+non_residential_property_secured_debt,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
+other_residential_property_secured_debt,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
+owned_land_secured_debt,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 pension_contributions_via_salary_sacrifice,1.0,1.059,1.127,1.205,1.261,1.308,1.337,1.365,1.396,1.431,1.431,1.431,1.431,1.431,1.431
 employment_income,1.0,1.059,1.127,1.205,1.261,1.308,1.337,1.365,1.396,1.431,1.431,1.431,1.431,1.431,1.431
 employment_income_before_lsr,1.0,1.059,1.127,1.205,1.261,1.308,1.337,1.365,1.396,1.431,1.431,1.431,1.431,1.431,1.431
@@ -59,6 +64,7 @@ pension_income,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384
 personal_pension_contributions,1.0,1.059,1.127,1.205,1.261,1.308,1.337,1.365,1.396,1.431,1.431,1.431,1.431,1.431,1.431
 petrol_spending,1.0,1.104,1.635,1.796,1.531,1.483,1.452,1.406,1.363,1.305,1.237,1.237,1.237,1.237,1.237
 private_pension_income,1.0,1.003,1.053,1.106,1.161,1.216,1.261,1.288,1.315,1.346,1.346,1.346,1.346,1.346,1.346
+private_pension_wealth,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 private_transfer_income,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 property_income,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 recreation_consumption,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
@@ -72,9 +78,11 @@ state_pension_reported,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1
 statutory_maternity_pay,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 statutory_paternity_pay,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 statutory_sick_pay,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
+stocks_and_shares_isa,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 student_loan_repayments,1.0,1.059,1.127,1.205,1.261,1.308,1.337,1.365,1.396,1.431,1.431,1.431,1.431,1.431,1.431
 sublet_income,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 transport_consumption,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
+unit_and_investment_trusts,1.0,1.0,1.092,1.147,1.19,1.223,1.258,1.297,1.34,1.384,1.384,1.384,1.384,1.384,1.384
 universal_credit_reported,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
 water_and_sewerage_charges,1.0,1.0,1.0,1.092,1.14,1.21,1.283,1.349,1.4,1.46,1.46,1.46,1.46,1.46,1.46
 winter_fuel_allowance_reported,1.0,1.04,1.144,1.209,1.237,1.277,1.301,1.327,1.353,1.38,1.38,1.38,1.38,1.38,1.38
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		Export `private_pension_wealth`, `directly_held_shares`, `unit_and_investment_trusts`, `stocks_and_shares_isa` and `cash_isa` from the Wealth and Assets Survey, and impute debt secured on non-main property and land (`other_residential_property_secured_debt`, `non_residential_property_secured_debt`, `owned_land_secured_debt`).
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		Split private pension wealth out of `corporate_wealth`, which now holds only shares, unit and investment trusts and stocks and shares ISAs, and add buy-to-let property to `other_residential_property_value`.