The Skills Shift Reshaping Insurance Hiring in 2026

The Skills Shift Reshaping Insurance Hiring in 2026

The structural economics of the insurance industry are forcing an abrupt recalibration of human capital requirements. Over the past three years, property and casualty (P&C) carriers have weathered elevated combined ratios driven by loss costs, severe convective storm volatility, and persistent inflation in replacement materials and medical care. Simultaneously, life and annuity carriers face yield curve recalibrations and shifting policyholder expectations around digital distribution.

In response, executive leadership teams have moved past the initial phase of experimental technology investments. The priority now is operational throughput and precision risk selection. Enterprise automation, proprietary generative models, real-time sensor streams, and cloud-native core systems (such as Guidewire Cloud and Duck Creek OnDemand) are no longer downstream IT initiatives—they are the foundational infrastructure driving underwriting margin and claims severity reduction.

This structural shift directly changes which skills command a premium in the labor market and which traditional skill sets are being commoditized. For Chief Human Resources Officers (CHROs), Vice Presidents of Talent Acquisition, and insurance CEOs, navigating this shift requires abandoning legacy hiring profiles. The historical playbook—hiring underwriters based primarily on regional broker relationships or claims adjusters based on physical inspection bandwidth—is failing to deliver margin performance.

This analysis details the exact capabilities declining in market value, the high-demand technical skill sets required for modern risk operations, regional talent dynamics, and the diagnostic frameworks required to recruit and retain the talent that will define market leaders through 2026 and beyond.

---

The Macro Pressure: Why Insurance Talent Markets Are Decoupling from Legacy Profiles

To understand the current hiring market, one must first look at the underwriting unit economics driving enterprise talent strategies. For decades, the primary lever for insurance growth was distribution capacity: more agents, broader broker networks, and manual underwriting desks capable of reviewing submission documents. Technical innovation was largely confined to central actuarial teams who updated rating engines on quarterly or annual release cycles.

That equilibrium has collapsed under three macro pressures:

1. Unprecedented Underwriting Margin Compression: Combined ratios across commercial property and personal auto reached historic highs between 2022 and 2024. Carriers can no longer rely on investment income from float to offset underwriting losses. Operational expense ratios must drop by 200 to 400 basis points across mid-market and enterprise carriers to maintain target ROEs.
2. Exponential Unstructured Data Volumes: A single commercial property submission no longer consists of a standard three-page ACORD form and two years of loss runs. It now includes satellite imagery, IoT sensor feeds, 500-page policy document histories, building permit records, and climate risk scores. Human underwriters operating without automated intake engines face a structural capacity ceiling.
3. Regulatory and Fair Lending Scrutiny: State insurance commissioners, led by enforcement bulletins in states like Colorado, California, and New York, are actively auditing algorithmic pricing and automated underwriting models for proxy discrimination. Underwriters and data teams must now defend the auditability and explainability of automated risk selections.

Benchmark: Mid-market commercial P&C carriers transitioning from manual intake to LLM-assisted triage report a 42% reduction in time-to-bind, while time-to-fill for specialized actuarial data engineers capable of building these automated pipelines has stretched to 75–110 days in tier-1 hubs.

Consequently, talent acquisition teams are finding that traditional job descriptions yield candidates who lack the technical proficiency required to operate within modern, automated workflows. Conversely, pure technology candidates from traditional SaaS backgrounds often lack the deep domain knowledge—such as policy wording nuances, statutory accounting principles, and reinsurance structures—required to build production-grade insurance platforms. The market now demands hybrid talent: technical professionals with deep domain literacy, and domain experts with operational technology capabilities.

---

Devalued Core: 4 Capabilities Facing Rapid Commoditization

As intelligent document processing (IDP), generative language models, and predictive analytics embed into core platforms, several traditional insurance functions are seeing declining market demand, lower relative compensation growth, and head-count reductions.

+-----------------------------------------------------------------------+
|                 CAPABILITIES FACING COMMODITIZATION                   |
+-----------------------------------------------------------------------+
| 1. Manual Intake & ACORD Transcription                                |
|    - Rule-based data entry and manual SOV processing.                 |
|                                                                       |
| 2. Low-Complexity Personal Lines Claims Triage                        |
|    - First-pass loss adjusting and simple property damage checks.     |
|                                                                       |
| 3. Deterministic Excel-Based Actuarial Modeling                       |
|    - Static GLMs and manual reserve calculation spreadsheets.         |
|                                                                       |
| 4. Scripted Policy Administration & Customer Support                  |
|    - Tier-1 call center support and basic endorsement processing.    |
+-----------------------------------------------------------------------+

1. Manual Intake, ACORD Form Extraction, and Loss Run Aggregation

Historically, insurance operations relied heavily on armies of underwriting assistants, offshore processing teams, and junior operations staff to receive PDF submissions via email, manually re-key ACORD 125/126/140 forms into policy administration systems, and standardize Schedule of Values (SOV) spreadsheets.

Today, custom-tuned OCR models combined with Retrieval-Augmented Generation (RAG) architectures automatically ingest, parse, validate, and enrich structured and unstructured submission data in seconds. Roles focused purely on data entry, manual document classification, and basic validation are experiencing rapid headcount reductions. Compensation for pure intake and processing roles has stagnated, with talent demand dropping over 35% across national P&C carriers since 2023.

2. First-Pass Adjusting for Low-Complexity Personal Lines Claims

In personal auto and simple homeowner claims, traditional desk adjusters spent hours collecting photos, validating coverage limits, cross-referencing repair shop estimates, and issuing initial payments.

The rise of computer vision models (integrated through platforms like Tractable, Mitchell, and CCC ONE) allows policyholders to upload smartphone photos of auto or property damage and receive an audited, automated repair estimate within minutes. Straight-through processing (STP) rates for simple auto glass and minor physical damage claims now routinely exceed 60% at top-tier carriers. Consequently, demand for entry-level desk adjusters without specialized fraud detection or complex litigation management skills has declined significantly.

3. Legacy SQL-Only Actuarial Modeling and Static Excel-Based Reserve Calculations

For decades, actuarial science relied heavily on static Generalized Linear Models (GLMs), legacy deterministic tools, and complex, manual Excel spreadsheets to establish reserves and construct rate filings.

While actuarial fundamentals remain critical, the market value of actuaries who rely exclusively on manual spreadsheet manipulation or basic SQL query building has dropped dramatically. Modern rating demands dynamic, real-time pricing models built in Python, R, and PySpark, capable of ingesting high-frequency external data. Actuarial profiles lacking programmatic machine learning exposure or real-time pipeline integration skills are increasingly restricted to back-office statutory reporting roles, commanding lower compensation trajectories than their predictive-modeling peers.

4. Templated Scripting in Policy Administration and Agent Support

Tier-1 customer service representatives and agency operations personnel traditionally spent their days answering routine broker calls regarding billing status, issuing certificates of insurance (COIs), processing standard mid-term policy endorsements, and answering policy coverage questions by reading static PDFs.

Enterprise conversational AI agents—trained directly on carrier policy forms, state-specific endorsements, and billing system APIs—now handle these conversational workflows via phone, chat, and broker portals without human intervention. The market demand for phone-based support staff who rely on static scripts has fallen off sharp cliffs, forcing talent leaders to restructure agency service desks around complex coverage dispute resolution and high-value broker relationship engineering.

---

The Rising Stack: 5 High-Demand Technical and Analytical Capabilities

As legacy administrative roles contract, carriers are competing fiercely for a new class of specialized talent. These professionals build, audit, scale, and manage the automated systems driving modern underwriting profit. Talent teams must reconfigure their sourcing pipelines to target these five critical skill profiles.

+-----------------------------------------------------------------------+
|                5 HIGH-DEMAND TECHNICAL CAPABILITIES                   |
+-----------------------------------------------------------------------+
| 1. Geospatial Risk Analytics & Climate Modeling                       |
|    - Moody's RMS, Verisk, HazardHub, Spatial SQL, Satellite Imagery   |
|                                                                       |
| 2. LLM Fine-Tuning & RAG Engineering for Coverage Policy              |
|    - LangChain, LlamaIndex, Vector DBs, Unstructured Policy Analysis  |
|                                                                       |
| 3. High-Throughput Telematics & Real-Time IoT Engineering             |
|    - Snowflake, Apache Kafka, Databricks, Sensor Stream Integration  |
|                                                                       |
| 4. Algorithmic Underwriting Audit & AI Compliance Governance          |
|    - Model Explainability (SHAP/LIME), NAIC Bulletins, Fair Lending   |
|                                                                       |
| 5. Parametric Product Design & Automated Smart Contracts              |
|    - Index-Based Underwriting, Oracle API Triggers, Swift Settlement  |
+-----------------------------------------------------------------------+

1. Geospatial Risk Analytics and Climate Modeling Engineering

With traditional historical weather patterns failing to predict recent wildfire, severe convective storm, and inland flooding losses, carriers are overhauling their property risk pricing. Modern underwriting relies on hyper-local, real-time spatial analysis.

Engineers and analysts who can manipulate spatial vector data, integrate satellite/imagery feeds (via Nearmap or EagleView), and build custom catastrophe risk overlays in tools like Moody's RMS, Verisk Touchstone, HazardHub, or open-source GIS libraries (QGIS, GeoPandas, Spatial SQL) are among the most sought-after hires in property insurance. These professionals bridge the gap between pure meteorology/climate science and commercial property underwriting rules.

• Target Experience: 3+ years in spatial data science, proficiency in Python (GeoPandas, Shapely), advanced Spatial SQL, and hands-on experience building custom API integrations into core underwriting engines.
• Impact on Hiring: Time-to-fill for senior geospatial engineers regularly exceeds 90 days due to intense competition from both legacy carriers and climate-tech insurtechs.

2. LLM Fine-Tuning and RAG Engineering for Unstructured Policy Analysis

Insurance is fundamentally an industry built on unstructured text: policy documents, manuscript endorsements, loss run histories, engineering inspection reports, and legal depositions. Carriers are deploying custom Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) frameworks to unlock this data.

However, generic off-the-shelf LLMs routinely hallucinate policy exclusions or misinterpret complex umbrella/excess liability structures. High-performing carriers are hiring AI platform engineers who specialize in fine-tuning open-source models, constructing hybrid keyword-vector search pipelines, and managing prompt orchestration frameworks (such as LangChain or LlamaIndex) integrated with enterprise vector databases (Pinecone, Milvus, Qdrant).

[ Unstructured Documents ] ---> [ Document Parsing Engine ]
(PDFs, Loss Runs, SOVs)           (Tesseract, Unstructured.io)
                                              |
                                              v
[ Vector Database ] <---------- [ Chunking & Embedding ]
(Pinecone, Qdrant)                (Text Embeddings)
        |
        v
[ RAG Pipeline / LLM ] <------- [ Hybrid Keyword/Vector Search ]
(LangChain, LlamaIndex)
        |
        v
[ Automated Underwriting Workbench ] ---> (Risk Selection & Coverage Triage)

• Target Experience: Proven track record of deploying domain-specific RAG applications in regulated environments, deep expertise in PyTorch or TensorFlow, vector DB architecture design, and direct experience with document-parsing frameworks (e.g., Unstructured.io, Tesseract, Azure AI Document Intelligence).
• Impact on Hiring: Candidates in this space often hold software engineering or computer science backgrounds; recruiting teams must sell them on the massive scale and rich data assets available within enterprise insurance balance sheets.

3. High-Throughput Telematics and Real-Time IoT Data Pipeline Engineering

In commercial auto fleet management, usage-based personal auto insurance (UBI), and commercial property water-leak monitoring, risk is no longer assessed once per year at renewal. It is priced and managed continuously based on real-time sensor streams.

Carriers need data engineers capable of architecting low-latency, highly available ingestion pipelines handling billions of telematics ping events daily. This requires mastery of distributed streaming architectures (Apache Kafka, AWS Kinesis), modern data lakes (Databricks, Snowflake), and real-time processing engines (PySpark, Flink).

Benchmark: The average cost-per-hire for a Principal Geospatial Risk Engineer in secondary tech hubs (e.g., Salt Lake City, Charlotte) ranges from $22,000 to $38,000, with base salaries jumping 18% year-over-year to $185,000–$225,000.

• Target Experience: Senior-level software/data engineering backgrounds with expertise in event-driven architecture, Snowflake Snowpark, Delta Lake, and high-volume API development.
• Impact on Hiring: These engineers are often recruited directly out of fintech, logistics, or big tech companies, requiring talent teams to offer competitive base pay alongside flexible remote work policies.

4. Algorithmic Underwriting Audit and AI Compliance Governance

As insurance carriers shift from human-driven underwriting desks to automated decision rules and machine learning rating models, regulatory oversight has intensified. State insurance departments require absolute transparency into why a risk was declined, why a surcharge was applied, or how a credit score proxy was constructed.

This has birthed a new, highly specialized cross-functional role: the Algorithmic Underwriting Compliance Officer. These professionals combine deep knowledge of insurance regulatory law (NAIC guidelines, state insurance codes) with practical knowledge of machine learning explainability techniques (SHAP values, LIME, counterfactual explanations). They audit proprietary models and vendor algorithms to ensure zero disparate impact and maintain audit trails for state examinations.

• Target Experience: Backgrounds in insurance regulatory law, quantitative model validation, compliance risk management, and comfortable interpreting Python-based ML explainability libraries.
• Impact on Hiring: These professionals are exceedingly rare. Carriers often form these teams by pairing senior regulatory compliance attorneys with senior data scientists in co-led governance pods.

5. Parametric Product Design and Automated Smart Contract Actuarial Science

Parametric insurance—where payouts are triggered automatically based on verifiable third-party data events (e.g., a hurricane reaching Category 3 windspeeds within a specific coordinate grid, or a river gauge hitting a specific flood stage)—is expanding rapidly across commercial property, agriculture, and business interruption lines.

Designing parametric products requires a blend of non-traditional actuarial science, financial engineering, and automated API integration. Parametric product managers and actuaries must model index correlations, define precise basis risk parameters, and connect policy engines to real-time oracle data sources (e.g., NOAA weather feeds, USGS earthquake sensors, flight radar systems).

• Target Experience: Fellowship/Associateship in CAS (FCAS/ACAS) or SOA with specialized experience in capital markets, cat bonds, weather derivatives, or alternative risk transfer (ART) mechanisms.
• Impact on Hiring: Highly competitive talent acquisition market centered in major financial and reinsurer hubs (New York, Chicago, Bermuda, London), commanding top-tier compensation structures.

---

Compensation Dynamics, Talent Clusters, and Regional Geographies

The geographic layout of insurance talent has fundamentally shifted. While historical powerhouses like Hartford, Connecticut, and Des Moines, Iowa, retain deep institutional knowledge in underwriting, policy administration, and legal compliance, modern technical insurance talent has clustered into distinct high-growth geographic corridors.

+-----------------------------------------------------------------------------------+
|                           REGIONAL INSURANCE TALENT HUBS                          |
+-----------------------------------------------------------------------------------+
|  [Hartford / Des Moines]    --> Core Underwriting, Legal, Regulatory & Actuarial  |
|  [Chicago / New York]       --> E&S, Complex Commercial, Specialty & Capital Markets|
|  [Charlotte / Atlanta]      --> Insurtech Data Pipelines, Telematics & Cloud Infra|
|  [Austin / Salt Lake City]  --> AI/LLM Engineering, RAG Architectures & Platforms  |
+-----------------------------------------------------------------------------------+

Key Regional Hubs and Talent Profiles:

• Hartford, CT & Des Moines, IA (The Legacy Cores): Dominant for traditional underwriting leadership, regulatory compliance, life/annuity operational scale, and casualty risk management. Sourcing AI or platform engineering profiles in these markets often requires compensating at top-of-market levels or offering full remote flexibility to tap broader talent pools.
• Chicago, IL & New York, NY (Commercial & Specialty Centers): Concentrated hubs for Excess & Surplus (E&S) lines, complex commercial property, catastrophe bond structuring, and broker management. Compensation here reflects high cost-of-living premiums, with aggressive base salaries and performance bonuses linked directly to loss ratio outcomes.
• Charlotte, NC & Atlanta, GA (Fintech & Infrastructure Corridors): Rapidly growing centers for cloud platform engineering, data pipeline architecture, and enterprise software integration. These markets offer an exceptional density of engineers transitioning out of financial services into insurance platforms.
• Austin, TX & Salt Lake City, UT (Technology Expansion Markets): Crucial sourcing grounds for AI platform engineers, spatial data scientists, and user experience (UX) designers building modern broker workbenches and automated claims platforms.

The following table outlines standard market compensation bands, typical time-to-fill, and 12-month attrition expectations across legacy versus modern insurance capabilities in 2026:

---

Diagnostic Framework: How Talent Teams Can Filter for Modern Skillsets

When recruiting for modern insurance technical and operational roles, talent acquisition teams cannot rely on traditional keyword searches or standard resume screens. Resume buzzwords like "AI exposure," "underwriting transformation," or "data-driven" are ubiquitous and often obscure a candidate's actual technical depth.

Recruitment leads must implement practical diagnostic screening frameworks to differentiate high-value technical practitioners from candidates whose experience is limited to high-level strategic oversight or legacy operations.

+-----------------------------------------------------------------------+
|                    TECHNICAL CANDIDATE EVALUATION                     |
+-----------------------------------------------------------------------+
|  RESUME RED FLAGS                      RESUME GREEN FLAGS             |
|  - "Experienced in AI tools"           - Named models (PyTorch, RAG)  |
|  - "Managed loss run processing"       - Built automated ingestion    |
|  - "Excel & SQL predictive models"     - Python pipelines & PySpark   |
|  - "Worked with IT on Guidewire"       - Custom REST API endpoints    |
+-----------------------------------------------------------------------+

1. Screening Underwriting Talent for Technical Capability

Modern commercial underwriters must act as "portfolio managers" who understand how to configure and query risk selection platforms, evaluate model confidence scores, and collaborate directly with data teams.

Resume Green Flags:

• Demonstrates clear experience using automated triage workbenches to manage portfolio loss ratios, rather than relying solely on individual account writing.
• Explicitly cites experience providing feedback to data science teams to refine risk models or adjust algorithmic rating rules.
• Cites comfort with advanced analytics platforms (e.g., Palantir Foundry, custom Snowflake workbenches, Spatial SQL dashboards).

Resume Red Flags:

• Heavy reliance on broker relationships as the sole competitive advantage, without mentioning portfolio optimization or automated intake usage.
• Career history focused exclusively on manual document review without exposure to digital submission triage or automated rating workflows.

Practical Interview Assessment Prompt:

"Describe a situation where our automated risk scoring engine rejected a high-premium commercial account that your broker partner insists is low risk. Walk us through how you audit the model’s data inputs, evaluate the spatial/loss history data, and decide whether to override the system recommendation. What metrics do you present to senior underwriting leadership to justify an override?"

2. Screening Data Engineering & AI Candidates for Domain Context

Hiring brilliant data scientists or AI engineers who do not understand insurance mechanics often results in technically impressive models that cannot be deployed operationally or survive regulatory examination.

Resume Green Flags:

• Highlights experience handling real-world, messy insurance data types (e.g., unstructured ACORD PDFs, loss run histories, multi-line policy schedules).
• Mentions compliance frameworks, model explainability techniques (SHAP/LIME), or state rate filing support.
• Direct experience integrating models into legacy policy administration systems (Guidewire, Duck Creek, Majesco) via REST APIs or streaming layers.

Resume Red Flags:

• Experience limited strictly to clean, Kaggle-style structured datasets without messy real-world ingestion challenges.
• Inability to explain how their predictive model impacts basic insurance financials (e.g., loss ratio, expense ratio, written vs. earned premium).

Practical Interview Assessment Prompt:

"We need to build an automated extraction and scoring pipeline for commercial property loss runs across 50 different loss run format variants from competing carriers. How do you design an ingestion and validation pipeline that handles missing data fields, detects fraudulent loss run entries, and alerts an underwriter when model confidence falls below 85%?"

---

Re-architecting the Talent Pipeline: Retraining vs. Market Acquisition

Facing long time-to-fill metrics and high market compensation premiums, forward-thinking insurance carriers are realizing that they cannot simply buy their way out of the skills deficit. Enterprise talent strategies must combine targeted external recruitment for core technical leadership with structured internal reskilling pathways for high-performing domain experts.

+-----------------------------------------------------------------------------------+
|                            BUILD VS. BUY DECISION TREE                            |
+-----------------------------------------------------------------------------------+
|  Role Profile               Strategy       Primary Execution Path                 |
|  -------------------------  ------------   -------------------------------------  |
|  AI/RAG Platform Engineers   BUY            Recruit from SaaS / Cloud Native      |
|  Spatial Data Engineers     BUY            Recruit from GIS / Climate Tech        |
|  Underwriting Technologists BUILD          Reskill Senior Underwriters (Tech Focus)|
|  Modern Claims Analysts     BUILD          Reskill Adjusters (Analytics & Fraud)  |
|  Actuarial Data Scientists  HYBRID         Internal Rotations + Specialized Hiring|
+-----------------------------------------------------------------------------------+

1. The Underwriting Technologist Pathway

Carriers can reskill experienced senior underwriters who possess deep policy and loss context into "Underwriting Technologists." These individuals serve as the crucial translation layer between business operations and engineering teams.

• The Curriculum: A 16-week internal fellowship covering basic Python for data analysis, SQL query development, system configuration logic in Guidewire/Duck Creek, model explainability principles, and agile product ownership.
• The Business Case: Transforming an existing $120,000-a-year commercial underwriter costs roughly $15,000 to $20,000 in dedicated training materials and backfill capacity. Sourcing an external product manager with insurance domain knowledge often requires $180,000+ base salaries plus executive search fees ($40,000+), while carrying a higher 12-month failure rate due to lack of deep policy knowledge.

2. The Modern Claims Analyst Pathway

Instead of laying off desk adjusters as simple claims become automated, progressive carriers transition top-performing adjusters into complex fraud investigators and claims data analysts.

• The Curriculum: Training in data visualization platforms (Tableau, PowerBI), advanced fraud detection tooling (Shift Technology, Friss), preliminary SQL, and litigation cost analytics.
• The Business Case: Reduces severance costs, protects institutional knowledge regarding complex coverage disputes, and builds a specialized internal unit capable of driving down claims severity—the primary cost driver across P&C balance sheets.

3. Sourcing Strategy for Non-Traditional External Talent

When external recruitment is non-negotiable—such as for specialized AI Platform Engineers or Real-Time Pipeline Architects—talent teams must overhaul their employer value proposition (EVP).

Tech talent often perceives insurance as slow-moving and burdened by legacy systems. Recruiter messaging must counter this perception directly by emphasizing:

1. Massive Technical Scale: Highlighting the enterprise data volumes available (e.g., multi-petabyte telematics streams, multi-decade loss datasets).
2. Direct Financial Impact: Demonstrating how an optimized risk-selection model directly impacts hundreds of millions of dollars in underwriting capital allocation.
3. Modern Technical Stacks: Showcasing cloud-native deployment environments (AWS, Azure, Databricks, Snowflake, Kubernetes) and modern toolchains rather than legacy languages.

---

Conclusion: The New Blueprint for Insurance Talent Leadership

The transformation of the insurance labor market is not a temporary trend driven by tech industry hype cycles. It is the direct result of a structural shift in how risk is underwritten, priced, and managed across P&C, Life, and Specialty markets.

As predictive algorithms, spatial analytics, LLM architectures, and real-time sensor streams take over operational workflows, the human capabilities that drive carrier performance are shifting rapidly. Legacy intake processing, simple claims desk adjusting, and manual spreadsheet modeling are giving way to geospatial risk engineering, algorithmic compliance governance, and RAG-driven policy analytics.

For executive talent leaders, staying competitive requires a clear, deliberate response:

• Audit existing workforce profiles to identify teams exposed to rapid automation and map out explicit internal reskilling pathways.
• Modernize sourcing channels by looking beyond traditional insurance markets into regional tech hubs like Charlotte, Atlanta, Austin, and Salt Lake City.
• Restructure interview frameworks to rigorously screen for technical depth in tech roles and operational tech fluency in underwriting and claims roles.
• Align compensation strategies with modern technology benchmarks rather than legacy regional insurance pay scales to secure top-tier engineering and analytical talent.

Carriers that realign their talent acquisition strategies with these analytical and technological realities will build a sustainable competitive advantage in underwriting profitability. Those that remain anchored to traditional job profiles will find themselves priced out of risk selection precision, saddled with inflated expense ratios, and unable to compete in an increasingly automated marketplace.

To navigate this transition successfully, forward-thinking CHROs and hiring managers partner with domain-specialized talent advisors who understand the precise intersection of insurance operations and modern technology infrastructure. TaaSFlow works closely with mid-market and enterprise carriers to evaluate workforce skill gaps, design targeted recruitment strategies, and deliver the technical, actuarial, and operational talent needed to secure underwriting margin and drive long-term balance sheet growth.