📊

Financial Services > Investment & Wealth Management

Portfolio Management

Systems for portfolio construction, performance measurement, attribution analysis, risk management, and rebalancing for asset managers and wealth advisors

₹53L Cr

India MF AUM

$100T+

Global AUM

Daily

Rebalancing Frequency

<5 min

Performance Lag

What Engineers Miss When They First Enter Portfolio Management

Portfolio management systems are where the fund manager's investment decisions become the positions held in a portfolio, and the engineering challenge is keeping an accurate, real-time view of those positions as a continuous stream of events — trades, corporate actions, income, expenses — changes them. A fund manager running an equity portfolio with positions across 50 stocks, several futures contracts, and a few options needs to see their current P&L against the day's benchmark within minutes of the market open. The position keeping system that provides this view must process every trade execution from the OMS (order management system), every exchange-confirmed allocation, and every corporate action — splits, bonuses, dividends, rights issues — that changes the economics of existing positions.

Performance attribution analysis is one of the most technically demanding outputs of a PMS. The standard is the Brinson-Hood-Beebower attribution model, which decomposes a portfolio's return relative to a benchmark into three components: allocation effect (did you overweight sectors that outperformed?), selection effect (within each sector, did you pick better stocks than the benchmark?), and interaction effect (the interaction between the two). Implementing this correctly requires reconstructing the portfolio's weights and the benchmark's weights at the start of each period, computing returns for each segment, and aggregating across the full portfolio. Getting this right matters: attribution reporting is what the fund manager presents to the investment committee to justify their decisions.

India's T+1 settlement and the growing adoption of algorithmic trading strategies have increased the real-time requirements on PMS. Legacy PMS platforms that ran end-of-day batch reconciliation — acceptable when settlement was T+2 and trading was manual — cannot provide the intraday position and risk views that modern fund managers need. The market for real-time PMS solutions that process exchange feeds directly and update positions intraday is a growth area, with both established vendors (SimCorp, Bloomberg AIM) and Indian startups (Kredent, Ffreedom) competing for institutional clients.

What Teams Actually Do Day To Day

  • 1Build the position keeping system: processing trade allocations from the OMS into holding records, applying exchange confirmations (NSDL/CDSL settlement confirmations for equities, clearing member statements for derivatives), handling corporate action adjustments (bonus shares, stock splits, rights issue allocations, dividend accruals), and maintaining an accurate real-time view of each fund's holdings.
  • 2Develop the P&L and performance calculation engine: computing unrealised P&L as the difference between current market value and average cost basis, computing realised P&L as the difference between sale proceeds and the lot's cost basis (FIFO, LIFO, or average cost depending on the fund's accounting policy), and computing time-weighted returns for performance reporting in compliance with GIPS standards.
  • 3Implement performance attribution: reconstructing portfolio and benchmark weights at the start of each attribution period, computing Brinson-Hood-Beebower attribution components for each sector and security, aggregating across the full portfolio to compute the total active return explanation, and generating the attribution report that the investment team uses to explain performance to clients.
  • 4Build the risk management module: computing VaR (Value at Risk) for the portfolio using historical simulation or parametric methods, tracking concentration risk (single security and sector limits from the fund's SEBI-mandated investment restrictions), monitoring derivative exposures (delta, gamma, vega for options positions), and generating limit breach alerts when positions approach regulatory or internal risk limits.
  • 5Operate the reconciliation process: comparing the PMS's position records with the custodian's records (DSP, HDFC Custodial, ICICI Securities Primary Dealership) and the exchange's settlement records daily, identifying and resolving breaks (discrepancies between the two sets of records) before they accumulate and affect NAV accuracy.

One End-to-End Flow: Fund Manager Buys 10,000 Shares of Infosys and the PMS Reflects the Position

A fund manager executes a purchase of Infosys shares through the trading desk. The trade goes through the OMS, is confirmed by the exchange, and the PMS reflects the new position in real time — with the full cost, cash impact, and P&L computed.

1

Trade order is placed and executed

The fund manager places a buy order for 10,000 Infosys shares through the OMS at a limit price. The OMS routes the order to the broker's execution desk or directly to the exchange via DMA. The order is filled at ₹1,820 per share (total ₹1.82 Cr). The execution report flows back to the OMS.

Systems Involved

OMS, broker DMA, exchange matching engine, FIX protocol execution report

Where It Usually Breaks

Partial fills — where only 7,000 of the 10,000 shares are executed before the trading day ends — require the OMS to split the order into a filled portion and an open portion. The PMS must handle partial allocations correctly and not record the full 10,000 shares until all are filled.

2

Trade is allocated to the fund and confirmed by the custodian

The OMS allocates the executed trade to the specific fund that ordered it. The allocation is transmitted to the custodian bank. The custodian verifies the trade against the settlement instructions and pre-funds the settlement obligation. NSE/BSE confirms the trade on T, and settlement — delivery of securities against payment — occurs on T+1.

Systems Involved

OMS allocation, custodian SWIFT messaging, exchange trade confirmation, T+1 settlement

Where It Usually Breaks

Allocation mismatches — where the broker's allocation record differs from the OMS's allocation record — cause settlement failures on T+1. The fund does not receive the securities, the custodian issues a buy-in notice, and the trade must be resolved manually with the broker at penalty rates.

3

PMS updates position and cost basis

Upon receiving the exchange confirmation, the PMS creates a new position record or adds to the existing Infosys holding: 10,000 shares at an average cost of ₹1,820. The cash balance is reduced by ₹1.82 Cr plus brokerage and STT. The fund's total equity exposure increases, and the portfolio's sector allocation dashboard updates to reflect the new Infosys position.

Systems Involved

PMS position keeping, cost basis calculation, cash balance update, sector allocation recalculation

Where It Usually Breaks

Corporate action timing errors — if a bonus share or stock split is recorded in the PMS on a different date than when the exchange actually credited the shares — cause temporary discrepancies between the PMS position and the custodian's account that must be identified and corrected in the daily reconciliation.

4

Performance attribution and risk metrics are updated

The intraday P&L engine marks the Infosys position at the current market price (using live exchange data feed) and computes the unrealised P&L. The risk engine recomputes the portfolio's concentration in the Technology sector and checks it against the fund's concentration limit. End of day, the attribution model updates the IT sector allocation and selection effects.

Systems Involved

Market data feed, real-time P&L engine, risk limit monitoring, end-of-day attribution batch

Where It Usually Breaks

Market data feed latency or stale prices during exchange circuit breaker halts cause the intraday P&L to show incorrect values. Risk teams must be aware of feed quality issues and treat risk reports from such periods as provisional.

Technology Architecture — How Portfolio Management Platforms Are Built

The diagram below reflects how production Portfolio Management systems are structured at scale — nine layers from client channels through edge security, API gateway, domain microservices, polyglot data stores, async event streaming, analytics, external partners, and cloud infrastructure. Solid arrows show synchronous REST/gRPC calls; dashed arrows show async event flows via Kafka or a message queue.

Portfolio Management — Enterprise Architecture ReferenceSolid arrows: synchronous calls (REST / gRPC) · Dashed arrows: async event flows (Kafka / Message Queue)CLIENTS & CHANNELSWeb SPAiOS / AndroidAdmin PortalPartner API3rd-Party WebhooksBatch / CronEDGE SECURITY & DELIVERYCDN (CloudFront / Akamai) · DDoS Shield · WAF (OWASP rules) · SSL/TLS Termination · Global Load Balancer (ALB / NLB)API GATEWAYKong / AWS API Gateway / NGINX / ApigeeRate Limiting · Routing · Versioning · Throttling · BFF PatternIDENTITY & ACCESSOAuth 2.0 · OpenID Connect · SAML 2.0JWT · RBAC · MFA · SSOCORE DOMAIN MICROSERVICES · REST / gRPC📋 Position Keeping & Holdi…Maintain book-of-record positionsReal-time mark-to-market valuationGET /api/v1/portfolios/{id}/positio…BlackRock Aladdin📈 Performance Measurement …Time-Weighted Return (TWR) calcul…Money-Weighted Return / XIRR for …GET /api/v1/portfolios/{id}/perform…FactSet PA⚠️ Risk Management & Analyt…Value at Risk (VaR) calculationExpected Shortfall (CVaR / ES)GET /api/v1/portfolios/{id}/varBlackRock Aladdin Risk⚖️ Portfolio Rebalancing En…Target allocation managementDrift detection vs tolerance bandsGET /api/v1/portfolios/{id}/driftiRebal (TD Ameritrade)Service Mesh: mTLS · Circuit Breaker (Resilience4j / Hystrix) · Service Discovery (Consul / Eureka) · Distributed Tracing (Jaeger)DATA PERSISTENCE · PolyglotPostgreSQL / Or…OLTPTimescaleDB / I…PrimaryRedis CacheCacheElasticsearchSearchS3 / BlobObjectASYNC MESSAGING & EVENTSApache Kafka / SQSPub/Sub · TopicsDead Letter QueueError HandlingStream ProcessorFlink / SparkANALYTICS & DATA PLATFORMData Warehouse (BigQuery / Snowflake / Redshift) · ETL/ELT (dbt / Airflow) · BI Tools (Tableau / Metabase) · ML Feature StoreEXTERNAL INTEGRATIONS & PARTNERSMarket Data Feed (Blo…Custodian BankDepository (CDSL/NSDL)OMSPosition Keeping Syst…Market DataPLATFORM: AWS / Azure · Kubernetes (EKS/AKS/GKE) · Docker · Helm · ArgoCD · CI/CD (GitHub Actions) · IaC (Terraform)OBSERVABILITY: ELK / Datadog · Prometheus / Grafana · Jaeger · PagerDutySECURITY: TLS 1.3 · Vault / KMS · SAST/DAST · SOC2 / ISO 27001Sync (REST / gRPC)Async (Kafka / Events)Each service owns its bounded context · CQRS & Event Sourcing where applicable · Polyglot persistence per domain

Industry Players & Real Applications

🇮🇳 Indian Companies

SBI Mutual Fund

Largest AMC India

Bloomberg AIM, Charles River

₹9L Cr+ AUM, largest in India

HDFC AMC

Private AMC

Simcorp Dimension, Custom

Top private AMC by AUM

Nippon India MF

Private AMC

Custom PMS, Oracle

Listed AMC with large AUM

Kotak Mahindra AMC

Private AMC

FactSet, Bloomberg, Custom

Strong institutional business

IIFL Wealth

Wealth Manager

Custom WMS, Bloomberg

HNI/UHNI wealth management

360 ONE (IIFL)

Wealth Platform

Proprietary + SS&C

Rebranded wealth + asset management

🌍 Global Companies

BlackRock (Aladdin)

USA

World's Largest Asset Manager

Aladdin Platform

$10T+ AUM, Aladdin used by 200+ firms

Vanguard

USA

Index Fund Pioneer

Internal + Broadridge

$8T+ AUM, inventor of index fund

Fidelity Investments

USA

Active + Passive AMC

Custom + Charles River

$4.5T+ AUM

PIMCO

USA

Fixed Income Specialist

Blackrock Aladdin + Custom

Bond market leader

Bridgewater

USA

Macro Hedge Fund

Fully Proprietary

World's largest hedge fund, principles-based

Man Group

UK

Quant Hedge Fund

Man AHL Proprietary

Largest publicly-listed hedge fund

🛠️ Enterprise Platform Vendors

BlackRock Aladdin

Portfolio Management, Risk Analytics, OMS

Industry benchmark — used by central banks

Charles River IMS

Portfolio Management, Compliance, OMS

Widely used by Indian AMCs

Simcorp Dimension

Investment Management, Fund Accounting

European market leader

SS&C Technologies

Advent Geneva, APX, Axys

Hedge fund and wealth management focus

FactSet

Portfolio Analytics, Performance, Risk

Analytics and data platform

Bloomberg PORT

Portfolio Risk & Analytics

Integrated with Bloomberg Terminal

Core Systems

These are the foundational systems that power Portfolio Management operations. Understanding these systems — what they do, how they integrate, and their APIs — is essential for anyone working in this domain.

Business Flows

Key Business Flows Every Developer Should Know.Business flows are where domain knowledge directly impacts code quality. Each flow represents a real business process that your code must correctly implement — including all the edge cases, failure modes, and regulatory requirements that aren't obvious from the happy path.

The detailed step-by-step breakdown of each flow — including the exact API calls, data entities, system handoffs, and failure handling — is covered below. Study these carefully. The difference between a developer who “knows the code” and one who “knows the domain” is exactly this: the domain-knowledgeable developer reads a flow and immediately spots the missing error handling, the missing audit log, the missing regulatory check.

Technology Stack

Real Industry Technology Stack — What Portfolio Management Teams Actually Use. Every technology choice in Portfolio Managementis driven by specific requirements — reliability, compliance, performance, or integration capabilities. Here's what you'll encounter on real projects and, more importantly, why these technologies were chosen.

The pattern across Portfolio Management is consistent: battle-tested backend frameworks for business logic, relational databases for transactional correctness, message brokers for event-driven workflows, and cloud platforms for infrastructure. Modern Portfolio Managementplatforms increasingly adopt containerisation (Docker, Kubernetes), CI/CD pipelines, and observability tools — the same DevOps practices you'd find at any modern tech company, just with stricter compliance requirements.

⚙️ backend

Java / Spring Boot

Core PMS engine, performance calculations, position keeping

Python

Risk models, attribution analysis, quantitative analytics, ML factor models

C++

Ultra-fast risk calculations, real-time pricing engines

Scala / Spark

Large-scale historical data processing, backtesting

🖥️ frontend

React / Next.js

Portfolio dashboards, investor portals, advisor workbenches

Angular

Internal fund manager trading workstations

D3.js / Highcharts

Performance charts, allocation visuals, risk heatmaps

🗄️ database

PostgreSQL / Oracle

Portfolio holdings, transactions, client data

TimescaleDB / InfluxDB

Time-series price data and performance history

Redis

Real-time P&L caching, live price streaming

Snowflake / Redshift

Analytics data warehouse, regulatory reporting

🔗 integration

Bloomberg B-PIPE

Real-time and historical market data

FIX Protocol

Order routing to brokers and exchanges

SWIFT ISO 20022

Settlement messages with custodians

Apache Kafka

Trade event streaming, position updates, market data

☁️ cloud

AWS

Most modern WealthTechs — S3 for data lake, EMR for analytics

Azure

Enterprise AMCs — Microsoft ecosystem integration

GCP BigQuery

ML workloads and large-scale portfolio analytics

Interview Questions

Q1.What is the difference between TWR and MWR/XIRR? When do you use each?

TWR (Time-Weighted Return) eliminates the effect of cash flows — used to evaluate fund manager performance. Calculated by chaining sub-period returns between cash flows. MWR/XIRR (Money-Weighted Return) accounts for timing of cash flows — used to measure actual investor experience. Example: fund manager returned 15% TWR, but investor who invested more before a drawdown might have 8% XIRR. Use TWR for manager comparison, XIRR for investor statements.

Q2.Explain Brinson attribution analysis.

Brinson-Hood-Beebower decomposes active return (portfolio vs benchmark) into: 1) Allocation Effect — did we over/underweight the right sectors? = (wp - wb) × (rb - R), 2) Selection Effect — did we pick better stocks within each sector? = wb × (rp - rb), 3) Interaction Effect — combined over/underweighting and selection. Sum of all three = active return. Used to understand if outperformance came from asset allocation or security selection skill.

Q3.How do you handle corporate actions in a portfolio system?

Corporate actions (dividends, splits, rights, mergers) must be applied accurately to maintain correct positions and cost basis. Process: 1) Receive corporate action notification (from exchange, data vendor), 2) Validate against holdings, 3) Apply: Stock split → multiply shares, divide price; Cash dividend → cash position increase, income posting; Bonus → add shares; Rights → offer subscription to investor. Need to handle ex-date vs record date. Must update cost basis, historical NAV for performance. Missing corporate actions = wrong performance calculations.

Q4.What is VaR and what are its limitations?

VaR (Value at Risk) is the maximum expected loss over a time period at a confidence level. Example: 1-day 95% VaR of ₹10L means 95% of days the loss won't exceed ₹10L. Methods: Historical simulation (actual past returns), Parametric (assume normal distribution), Monte Carlo. Limitations: 1) Doesn't tell you the loss beyond the threshold (tail risk), 2) Assumes normal distribution — underestimates fat tails, 3) Based on historical data — doesn't capture regime changes, 4) Not additive across desks simply. CVaR/Expected Shortfall addresses the tail risk gap.

Q5.How does T+1 settlement impact portfolio systems?

India's T+1 settlement means trades must settle (money and securities exchanged) by next business day. Portfolio system impacts: 1) Position updates must be real-time — can't wait for overnight batch, 2) Cash availability checks must consider T+1 payables, 3) Corporate action processing tighter timelines, 4) Custodian reconciliation must happen same day, 5) NAV calculation for MFs must account for T+1 settled positions. Systems need STP (Straight-Through Processing) with no manual intervention.

Glossary & Key Terms

AUM

Assets Under Management — total market value of funds managed

TWR

Time-Weighted Return — portfolio return eliminating effect of investor cash flows

XIRR

Extended Internal Rate of Return — investor return accounting for cash flow timing

VaR

Value at Risk — maximum expected loss at a given confidence level

Attribution

Decomposition of portfolio returns into allocation, selection, and interaction effects

Benchmark

Reference index (e.g., Nifty 50) to compare portfolio performance against

Tracking Error

Standard deviation of active returns (portfolio vs benchmark)

Alpha

Return in excess of benchmark — measure of manager skill

Beta

Sensitivity of portfolio returns to market movements

Rebalancing

Restoring portfolio to target allocation after price movements cause drift

GIPS

Global Investment Performance Standards — ethical standards for calculating and presenting returns

Sharpe Ratio

Risk-adjusted return = (Portfolio Return − Risk-Free Rate) / Portfolio Std Dev