Re: Isn't a lot of the trading based on algorithm's?


[ Follow Ups ] [ Post Follow Up ] [ UCLA Open Forum ]

Posted by confused442 on December 05, 2025 at 19:07:49

In Reply to: Isn't a lot of the trading based on algorithm's? posted by SagoBob on December 05, 2025 at 16:57:23

I’m not an expert, but here’s how I understand things.

> “Then all it would take for some computer-based trading schemes to scream ‘SELL’ and millions of shares will trade downward in nanoseconds.”

Sort of. About 85% of trading is done by algorithms, but not all bots act the same. Each algorithm is different, and modern trading systems adapt to market conditions. For example, if there’s a major sell-off, bots can switch to a different strategy to avoid compounding the drop. After Black Monday in 1987, regulators implemented circuit breakers: if the market falls beyond a certain threshold, trading pauses. This gives investors—and the bots—a moment to recalibrate and make sure the right strategies are in place.

> “And I read that Google has developed some products that performance-wise perform as well or better than Nvidia’s.”

Here, I’m less certain, but my understanding is this: Google produces application-specific chips, optimized for certain tasks, whereas Nvidia makes general-purpose GPUs. Google’s chips can outperform Nvidia on very specific benchmarks, but Nvidia’s hardware is more flexible overall. Either way, there’s plenty of market space for both, and likely others, to prosper.

What should be keeping us awake is China’s AI advantage. On some metrics, Chinese AI is already better than ours—and it’s much cheaper, sometimes multiple times cheaper.

Why? China has heavily invested in low-cost renewable energy. Their electricity is far cheaper than in the U.S.—more than twice as cheap today, and it could soon be three times cheaper. That means the cost of running AI and high-performance compute in China will soon be three times lower than in the U.S., giving them a substantial structural advantage.


Follow Ups:



Post a Followup

Name:
Email:
Password:

Subject:

Comments:

Optional Link URL:
Link Title:
Optional Image URL:


[ Follow Ups ] [ Post Follow Up ] [ UCLA Open Forum ]