Management Action: Implementing Dividend-Based Stress Responses in SAS Stress Testing

[#1]

One of the most important characteristics of modern stress testing is that institutions are no longer passive recipients of economic shocks. Banks respond strategically to stress through management actions such as reducing dividends, conserving liquidity, tightening lending standards, deleveraging portfolios, and restructuring balance sheets. Consequently, credible stress testing frameworks must not only model economic deterioration but also model how management behavior changes under stress conditions.

Among all available management interventions, dividend reduction remains one of the fastest and most operationally effective capital preservation mechanisms available to financial institutions. Reducing shareholder distributions immediately increases retained earnings and stabilizes projected capital positions without requiring external capital issuance or significant operational restructuring.

In this post we examine how management judgment can be operationalized inside a SAS-based stress testing framework through dividend-based management actions. Using SAS implementation logic and a sample set of Profit-and-Loss and Balance Sheet input, output, and delta projections, the analysis demonstrates how dividend management actions propagate through financial statements and materially improve solvency outcomes under adverse conditions.

Econometric Foundations of Stress Testing

Traditional stress testing frameworks generally rely on econometric relationships that forecast future financial variables using macroeconomic drivers and historical (autoregressive) persistence. Under normal operating conditions, such autoregressive behavior may reasonably approximate future outcomes. However, during severe stress environments, management often intentionally breaks from historical behavior patterns.

A bank that historically maintained stable shareholder distributions may sharply reduce dividends under stress in order to preserve capital. Consequently, stress testing systems require explicit management overlays capable of overriding historical persistence and embedding strategic policy decisions directly into forecast calculations.

SAS Implementation of Dividend Management Action

Management Actions are implemented through the following SAS options presented in the Management Action Task in the SAS Stress Testing Cycle:

Select any image to see a larger version.
Mobile users: To view the images, select the "Full" version at the bottom of the page.

The implementation logic is encapsulated within the model ST Financial Statement Management Action - <yyyy.mm>. In this example, we trace how a stress payout of 5 cents by management affects the Profit & Loss statement and Balance Sheet figures.

The management’s chosen dividend-per-share assumption is captured in the first few lines of the model through the variable DPS:

%let DPS=${Params.DPS};
data &paramStaticArray. ;
set &paramStaticArray.;
if param_name="PL_VALUE|||Dividend_per_Share___L1" then param_value=0;
if param_name="PL_VALUE|||Dividend_per_Share___constant" then param_value=&DPS./100;
run;

The stress payout of 5 cents is captured through the transformation:

param_value=&DPS./100;

and this converts the input into a decimal payout value of:

Dividend Per Share = 0.05

This represents a severe dividend suppression strategy designed to preserve capital under adverse conditions. The second part of the code removes historical payout persistence:

if param_name="PL_VALUE|||Dividend_per_Share___L1" then param_value=0;

Under normal forecasting conditions, dividend behavior might follow a relationship such as:

where current dividends depend partly on prior-period payouts.

By forcing the lag coefficient to zero, the model intentionally disconnects future dividend projections from historical payout inertia. This is a critical conceptual feature because it demonstrates how management judgment overrides statistical continuation during periods of stress.

%let var=PL_VALUE|||Dividend_per_Share;
 
BEP["&var.",t]=
AUX["&var.___constant"]
+ AUX["&var.___GDPRatioL0"] * scenario["gdp_Ratio",t]
+ AUX["&var.___GDPRatioL1"] * scenario["gdp_Ratio",t-1]
+ AUX["&var.___L1"] * BEP["&var.",t-1];

This corresponds to the structure:

However, because the lag parameter has been eliminated and the constant term is externally controlled by management, the dividend forecast becomes primarily policy-driven rather than statistically extrapolated. The model therefore transitions from passive forecasting into active strategic execution.

Profit-and-Loss Impact of Dividend Reduction

The impact of this management action becomes visible through the Profit-and-Loss input, output, and delta tables. Under the adverse scenario before management intervention, Dividend_per_Share remains stable at 50 cents across the forecast horizon from 2023 through 2026. This represents the stressed baseline assumption without capital conservation measures.

Profit & Loss Input Table (Before Management Action)

Following the management action, the output table shows Dividend_per_Share declining from 50 cents to 5 cents beginning in 2024:

Profit & Loss Output Table (After Management Action)

The corresponding delta table captures the difference between output and input projections:

Delta = Output − Input

The resulting dividend delta equals: 0.05 − 0.50 = −0.45 for 2024 through 2026.

Importantly, this negative delta does not imply negative dividend payouts. Rather, it indicates that management reduced projected dividends by 45 cents relative to the original stressed projection. This interpretation aligns fully with enterprise stress testing practice, where delta analysis isolates the incremental effect of management intervention.

Profit & Loss Delta Table

The Profit-and-Loss tables also reveal one of the most analytically significant findings of the exercise. Before management action, projected Profit_After_Distribution under the adverse scenario deteriorates materially. In 2024, Profit_After_Distribution falls to negative 356,414 (in the P&L Input table). Following the dividend reduction management action, however, Profit_After_Distribution improves to 993,586 (in the P&L Output table). The delta table shows the exact improvement:

+993,586 – (-356,414) = +1,350,000

This improvement precisely matches the preserved dividend outflow. With 3 million shares outstanding and a 45-cent reduction in Dividend-per-Share, the preserved capital equals:

3,000,000 × 0.45 = 1,350,000

The institution therefore preserves approximately $1.35 million annually through reduced shareholder distributions. The same improvement continues through the remainder of the stress horizon, with Profit_After_Distribution improving by $1.35 million in each forecast period.

One of the most important observations is that the management action does not improve operating profitability itself. The delta tables show no changes in Profit_Before_Income_Tax, Profit_After_Income_Tax, Net_Interest_Income, or Net_Operating_Income.

This demonstrates that the management action is not generating new earnings. Instead, it is preserving existing capital by reducing cash outflows to shareholders. This distinction is highly important from a regulatory and analytical perspective because it shows that management action improves solvency through capital conservation rather than optimistic revenue assumptions.

Balance Sheet Stabilization Under Stress

The transmission mechanism continues into the balance sheet projections. Under the adverse scenario before management action, projected balance sheet values deteriorate significantly. Projected assets decline to approximately $98.6 million in 2024, while projected equity weakens materially under stress conditions. The balance sheet input values under the adverse scenario show assets at approximately $98.62 million, liabilities at negative $18.95 million, and equity at negative $79.68 million in 2024.

Balance Sheet Input Table (Before Management Action)

Following the management action, the balance sheet output table shows improved projected asset values of approximately $99.97 million in 2024. The balance sheet delta table isolates the direct effect of the dividend reduction, showing an asset increase of $1.35 million and an equity movement of negative $1.35 million, while liabilities remain unchanged.

Balance Sheet Output Table (After Management Action)

The same cumulative effect continues through the forecast horizon. By 2025, the asset delta increases to $2.70 million, and by 2026 the cumulative asset preservation effect reaches $4.05 million. The corresponding equity delta values move by the same magnitude. The sign convention used by the stress testing framework stores equity as a balancing negative value. Economically, however, the increasing magnitude reflects improving retained capital preservation.

Balance Sheet Delta Table

The balance sheet movement exactly mirrors the preserved dividend amount accumulated across forecast horizons. By suppressing shareholder payouts, cash remains within the institution and strengthens the projected asset base. Conceptually, the relationship can be summarized as:

Reduced Dividend Outflow → Higher Retained Cash → Stronger Asset and Equity Position

This demonstrates the full propagation chain of management action through the financial statements:

Macroeconomic Stress → Profitability Pressure → Dividend Reduction → Retained Earnings

Preservation → Balance Sheet Stabilization

Balance Sheet Reconciliation and Accounting Consistency

The balance sheet reconciliation process is maintained through additional SAS balancing logic in the model. For example, in the following code, the first statement will take whatever Deferred Tax Expense was projected in the Income Statement and create the corresponding Deferred Tax Asset on the Balance Sheet.

BEP['BS_VALUE|||ASSET|||TAXES|||DEFERRED',t]=
-BEP['PL_VALUE|||TAXES|||EXPENSE|||DEFERRED',t];

BEP['BS_VALUE|||LIABILITY|||TAXES|||DIRECT',t]=
BEP['PL_VALUE|||TAXES|||EXPENSE|||DIRECT',t];

%apply_balancing_rules;

Similarly, the second statement above will take whatever Direct Tax Expense was projected in the Income Statement, create the corresponding Tax Liability on the Balance Sheet.

The real balancing happens when the macro %apply_balancing_rules is executed after all projections have been calculated. This macro adjusts designated balancing accounts to ensure that the projected Balance Sheet remains internally consistent and satisfies the accounting identity Assets = Liabilities + Equity.

The most important contribution of this framework is the use of delta analysis to isolate the incremental effect of management action. Without delta analysis, analysts observe only stressed projections. With delta analysis, they can quantify how much capital was preserved, how much solvency improved, how effective the management intervention was, and how management behavior alters stress outcomes. The result is a fully integrated enterprise stress testing architecture capable of modeling how management intervention propagates dynamically through institutional solvency outcomes.

Additional Information

For more information on SAS Stress Testing visit the software information page here. For more information on curated learnings paths on SAS Solutions and SAS Viya, visit the SAS Training page. You can also browse the catalog of SAS courses here.

Find more articles from SAS Global Enablement and Learning here.