The live online tests for the 2027 Test Series begin on Saturday, 1 August 2026. Tests are held every Saturday from 5:00 pm IST.
There are 30 live tests in the cycle. They can only be taken on schedule and are not available afterwards, so anyone joining later starts from the next scheduled test.
The full list of what the Test Series includes, the fee, and the registration link are on the Test Series page. Registration stays open through the cycle. If you are enrolled in the Preparatory Program, the Test Series is included and you do not need to register separately.
Registration is open for Quantitative Foundations for Economics, a weekly live course beginning Sunday, 30 August 2026. Sessions run for 1.5 hours every Sunday, from 8:30 to 10:00 AM IST, for about 45 hours of lecture time across roughly 30 sessions.
The course covers logic and sets, real analysis, linear algebra, convexity and optimization, probability and statistics, econometrics, and R programming. Sessions are live only and are not recorded. The fee is ₹40,000.
Registration is open for Seeing Probability: The Law of Large Numbers, the CLT, and Python — a two-day live workshop on 25 July and 1 August 2026. It uses Python simulations to make two central results in probability visible: what convergence looks like when it holds, and what happens when its assumptions are violated.
Python is used as a tool for exploring the mathematics, not as a separate subject. No prior programming experience is required, though you should be comfortable with undergraduate-level probability. All code runs in Google Colab, so nothing needs to be installed.
Dates: 25 July and 1 August 2026 — two 3-hour live sessions, one week apart
Econschool now has a discussion forum at forum.econschool.in. It’s a place to ask questions, work through problems, and discuss economics — open to students at any level, not only those preparing for entrance exams.
Post your questions in Questions & Problems, and use General for study advice, introductions, and wider discussion. Announcements will carry notices like this one. The forum supports LaTeX, so mathematical questions and answers render properly.
Everyone is welcome to read; posting requires a free account. See you there.
Registration is open for the Preparatory Program for the 2027 intake of the ISI MS/PhD entrance in Quantitative Economics. Live online lectures begin on Sunday, 14 June 2026, and run about 2.5 hours each Sunday — roughly 100 hours of instruction in total, covering Microeconomics, Macroeconomics, Mathematics, Probability, Statistics, and Econometrics. The fee is Rs. 80,000 and includes the integrated Test Series.
Full details — syllabus, schedule, eligibility, and how to register — are on the Preparatory Program page.
This tool shows UMP and EMP when preferences are satiated: there is a bliss point \( (\delta_x, \delta_y) \) where utility peaks, and utility falls as the bundle moves away from it. Preferences are therefore non-monotone — more is not always better (local nonsatiation fails). When the bliss point is affordable, the consumer always buys it, whatever their income, and leaves unspent any income beyond its cost, so the budget need not bind and Walras’ law fails. These are the definitions of UMP and EMP.
\[ \text{(UMP)}\qquad \max_{(x,y) \in \mathbb{R}^2_+}\; u(x,y) \quad\text{s.t.}\quad p_x x + p_y y \le M \]
\[ \text{(EMP)}\qquad \min_{(x,y) \in \mathbb{R}^2_+}\; \bigl(p_x x + p_y y\bigr) \quad\text{s.t.}\quad u(x,y) \ge \bar u \]
Choose how distance is measured and a problem, then move the bliss point, prices, and income, and watch the optimal bundle respond.
What to look for
Lower the prices or raise income until the bliss point is affordable. Does the consumer still always spend all income?
Make the bliss point unaffordable. Where does the optimum sit relative to the budget line and the bliss point?
Switch the distance measure between Euclidean and taxicab, and observe how the optimum changes.
This tool shows UMP and EMP when preferences are convex, i.e., when the upper-contour set \( \{\, (x,y)\in\mathbb{R}^2_+ : u(x,y) \ge \bar u \,\} \) is convex. Each of the four utilities considered in the tool has a convex upper-contour set, so the set of optimal bundles is always convex. These are the definitions of UMP and EMP.
\[ \text{(UMP)}\qquad \max_{(x,y) \in \mathbb{R}^2_+}\; u(x,y) \quad\text{s.t.}\quad p_x x + p_y y \le M \]
\[ \text{(EMP)}\qquad \min_{(x,y) \in \mathbb{R}^2_+}\; \bigl(p_x x + p_y y\bigr) \quad\text{s.t.}\quad u(x,y) \ge \bar u \]
Choose a utility function and a problem, then move the prices and income and watch the optimal bundle respond.
What to look for
Find the interior tangency on the smooth arm. Does the MRS equal the price ratio there?
Move the price ratio until the optimum sits at a kink in EMP. Observe the range of price ratios for which it remains optimal.
Slide a price slowly. Does demand jump, or move continuously?
Switch between UMP and EMP, and observe the nature of the optimal bundle.
This tool shows UMP and EMP when preferences are non-convex, i.e., when the upper-contour set \( \{\, (x,y)\in\mathbb{R}^2_+ : u(x,y) \ge \bar u \,\} \) is non-convex. Each of the four utilities considered in the tool has a non-convex upper-contour set. These are the definitions of UMP and EMP.
\[ \text{(UMP)}\qquad \max_{(x,y) \in \mathbb{R}^2_+}\; u(x,y) \quad\text{s.t.}\quad p_x x + p_y y \le M \]
\[ \text{(EMP)}\qquad \min_{(x,y) \in \mathbb{R}^2_+}\; \bigl(p_x x + p_yy\bigr) \quad\text{s.t.}\quad u(x,y) \ge \bar u \]
Choose a utility function and a problem, then move the prices and income and watch the optimal bundle respond.
What to look for
Find the tangency point, if it exists. Is it the best bundle?
Slide a price through the knife-edge. Does demand move smoothly, or jump?
Switch between UMP and EMP, and observe the nature of the optimal bundle.
A two-day live workshop · 25 July & 1 August 2026 · 10:00 AM – 2:00 PM IST (10:00–11:30, break, 12:30–2:00)
This workshop uses Python to make two central ideas in probability — the Law of Large Numbers and the Central Limit Theorem — easier to see and understand. Students who already know probability in theory will use simulations to observe what convergence actually looks like, when it does not occur as expected, and why it matters.
Python is used as a tool to explore the mathematics, not as a separate subject. No prior programming experience is required, but students should be familiar with undergraduate-level probability.
What you’ll do
Build sampling and simulation from first principles in NumPy
Compute expectations and variances by Monte Carlo, and understand the statistical properties of the estimates themselves
Watch the LLN converge and watch it fail when its assumptions are violated
See the CLT emerge from different distributions, and see where it breaks
Work on a problem set between the two sessions, with selected submissions discussed at the start of Day 2
Format
Two 3-hour live sessions, one week apart. Mode is: Online via Google Meet
Limited to 20 students. All code runs in Google Colab — no installation required.
Firm \( 1 \) has constant marginal cost \( c_1 = 1 \) and firm \( 2 \) has constant marginal cost \( c_2 = 2 \). The firms compete in quantities, choosing \(q_1\) and \(q_2\) simultaneously.
Find the Nash (Cournot) equilibrium quantities \( (q_1^*, q_2^*) \), the market price, and each firm’s profit.
Hint
Each firm maximizes its own profit taking the rival’s quantity as given. Write firm \( i \)’s profit, find a best-response function in \( q_j \). Two best responses, two unknowns.
Solution
For \( q = q_1+q_2 \in [0,12] \), inverse demand is \( p = 12-q \). Firm \( 1 \) chooses \( q_1 \in [0, 12-q_2] \) to maximize
\[ \pi_1 = \big(12-(q_1 + q_2)-1\big)\, q_1. \]
when \( q_2 < 12 \).
The first-order condition gives the best-response for the case \( q_2 < 11 \):
