Posts

Showing posts from November 18, 2025

Bajaj Finance Shares Fall 8% as Asset Quality Weakens in Q2

  A Sudden Slide in the Stock Shares of Bajaj Finance slipped sharply on November 11, tumbling over 8% intraday to hit ₹997 on the NSE. Even though the company posted healthy quarterly numbers, the Street reacted negatively to the worsening asset quality , pulling the stock briefly below the ₹1,000 mark. Last seen, the stock recovered slightly but was still down 7.54% at ₹1,003.20 . Why the Market Reacted: Asset Quality Concerns GNPAs and NNPAs Rise The company reported a clear deterioration in its asset book: Gross NPAs: 1.24% (up from 1.06% last year) Net NPAs: 0.6% (up from 0.46%) For an NBFC of Bajaj Finance’s scale, even small increases in NPAs signal rising stress. Capital Adequacy Still Healthy The capital adequacy ratio remained strong at 21.23% , giving the company enough buffer for future growth and risk absorption. Q2 Performance: Strong Numbers, Soft Sentiment Profit & Income Growth Despite NPA concerns, Bajaj Finance delivered solid financials: Ne...

The Rise of Self-Driving Cars and the Power of Machine Learning

  A Tragic Crash in Florida In May 2016, a Tesla in Florida hit a turning truck. The Tesla was behind the truck. The car did not slow down. Its roof was torn open in the impact. The Tesla kept moving and finally stopped after hitting another object. The truck driver was safe. The Tesla driver did not survive. What Investigators Found The autopilot was ON. The driver trusted the system. When the truck made a wrong turn, the autopilot failed to react in time. This incident showed both the power and the limits of self-driving tech. How It All Started ALV — The First Attempt In the 1980s, DARPA funded a project called the Autonomous Land Vehicle (ALV) . A simple system was used: Green pixels = grass Blue pixels = sky Brown pixels = road This helped the vehicle understand the path. It worked, but only at 3 km/h and only on simple dirt roads. When colours changed, the system failed. The world was too complicated for fixed rules. The Breakthrough: Neural...