Live data from Hacker News

They say that stocks go down during the day and up at night

statmodeling.stat.columbia.edu

61–70 of 154 posts

Re: They say that stocks go down during the day and up at night

#61

Since the article is not very informative, and it's a real rabbit hole to try to track this stuff down across all the linked articles etc. and as I don't even really care about stocks and the market, I am only left with one question which I did not find answered anywhere yet - what exactly does day and night mean in the context of the whole world trading?

> I am only left with one question which I did not find answered anywhere yet - what exactly does day and night mean in the context of the whole world trading? While we're at it, I have a related question: why do the exchanges even "open" and "close"? Surely in our globalized digital economy, it's not just "day" and "night" that are meaningless, but the very concept of "opening hours" itself.

I don't know if it's the reason, but I understand that a lot of market-moving news is released after close so that participants have time to digest it and decide whether their positions still make sense. The next day's opening auction will deal with the increased volatility better (see the story about the NYSE "forgetting" to hold the opening auction and the resulting chaos)

Re: They say that stocks go down during the day and up at night

#62

There are several explanations for this phenomena: * The largest moves in price occur after quarterly earnings announcements, which are released after-hours. Same with other material announcements. * Prices are more volatile after-hours because there are fewer market participants. Because the order book is smaller, the same sized trade will have a larger effect on price after-hours compared to during trading hours. *…

Add to all of these that the return is uninvestable, so the market cannot eliminate the arbitrage.

1016x over 30 years is a gain of less than 0.001% per trading day. You would have spent much more than this through fees and market impact to buy and sell the stock each day, especially at any real volume.

The market is full of uninvestable or tiny capacity trades like this. A reliable pattern gets created by trading flows like the ones described in other replies, and it doesn't get arbed out because it is uninvestable or has tiny capacity.

Re: They say that stocks go down during the day and up at night

#65

There are several explanations for this phenomena: * The largest moves in price occur after quarterly earnings announcements, which are released after-hours. Same with other material announcements. * Prices are more volatile after-hours because there are fewer market participants. Because the order book is smaller, the same sized trade will have a larger effect on price after-hours compared to during trading hours. *…

> holding overnight confers more risk, which will be rewarded by the market with higher returns.

That’s really not how risk works.

Re: They say that stocks go down during the day and up at night

#66

There are several explanations for this phenomena: * The largest moves in price occur after quarterly earnings announcements, which are released after-hours. Same with other material announcements. * Prices are more volatile after-hours because there are fewer market participants. Because the order book is smaller, the same sized trade will have a larger effect on price after-hours compared to during trading hours. *…

> holding overnight confers more risk, which will be rewarded by the market with higher returns. That’s really not how risk works.

Assuming an efficient market, that's exactly how it works - if there's more risk investors will require more reward to compensate.

Re: They say that stocks go down during the day and up at night

#67

There are several explanations for this phenomena: * The largest moves in price occur after quarterly earnings announcements, which are released after-hours. Same with other material announcements. * Prices are more volatile after-hours because there are fewer market participants. Because the order book is smaller, the same sized trade will have a larger effect on price after-hours compared to during trading hours. *…

Add to all of these that the return is uninvestable, so the market cannot eliminate the arbitrage. 1016x over 30 years is a gain of less than 0.001% per trading day. You would have spent much more than this through fees and market impact to buy and sell the stock each day, especially at any real volume. The market is full of uninvestable or tiny capacity trades like this. A reliable pattern gets created by trading fl…

Imagine how many HFT algorithms are out there trying to exploit every micro trend to the extreme. I am not in that game, but I can imagine all the things I would try, and I know there are really smart people and obviously a ton of money invested in it. Trying to find something simple that consistently makes money for a long time would be nearly impossible because the patterns would be found and exploited by many until they scrubbed out any potential gains. (But then again if someone did find it they wouldn't share). So you need to get novel and pour all sorts of external data into your models to have any edge. Something this simple just wouldn't be a thing. But it is still interesting to understand.

Re: They say that stocks go down during the day and up at night

#68
post #66

Earlier quoted context omitted.

> holding overnight confers more risk, which will be rewarded by the market with higher returns. That’s really not how risk works.

Assuming an efficient market, that's exactly how it works - if there's more risk investors will require more reward to compensate.

That’s a massive assumption. How efficient and rational does the stock market seem to you? With most of it driven by index funds and 401k contributions and even wall street traders jumping on bandwagons.

Risk reduces alpha. Risk doesn’t increase alpha. Yes obviously in a perfectly rational environment there would be a “need” to reward necessary risk but the market definitely does not need to in this case. More than enough idiots will hold their positions overnight without any upside.

Re: They say that stocks go down during the day and up at night

#69
post #10

Earlier quoted context omitted.

As someone who doesn't know much about stocks and trading - is it normal to keep stocks for only evening or evening -> morning, what about morning -> morning, or even longer timespans? Or is this the difference between trading and investing?

I believe when people talk about trading and investing the main difference is the time horizon. See Swing Trading, Day Trading etc. for shorter-term time-horizons. Keeping stocks only intra-day would be Day Trading (generally). There are lots of people on the internet who say they can 'teach' you day-trading. Don't do it unless you have a very high risk tolerance, i.e. are willing to lose it all and walk away. Even t…

> There are lots of people on the internet who say they can 'teach' you day-trading. Don't do it unless you have a very high risk tolerance,

Yeah I know, watched a ton of videos documenting the shady things these finance-gurus have done over the years. For a couple of months I was involved in a signal-trading group (I knew them personally), but I never had a good feeling about it and pulled my money without loosing much of it.

After playing around with crypto I've come to the conclusion that a savings and retirements account is probably the better solution for me, even though the interest is laughable. At least the money is safe as long as the global financial system doesn't totally collapse.

In the end I don't care enough about having more money to venture into investing etc., and I'm too scared of being conned by someone.

Re: They say that stocks go down during the day and up at night

#70

Earlier quoted context omitted.

Add to all of these that the return is uninvestable, so the market cannot eliminate the arbitrage. 1016x over 30 years is a gain of less than 0.001% per trading day. You would have spent much more than this through fees and market impact to buy and sell the stock each day, especially at any real volume. The market is full of uninvestable or tiny capacity trades like this. A reliable pattern gets created by trading fl…

Imagine how many HFT algorithms are out there trying to exploit every micro trend to the extreme. I am not in that game, but I can imagine all the things I would try, and I know there are really smart people and obviously a ton of money invested in it. Trying to find something simple that consistently makes money for a long time would be nearly impossible because the patterns would be found and exploited by many unti…

Algo trading teams are indeed research shops that have to continually replace their strategies. Often times success is less about the cleverness of a prediction than clever engineering (technical or contractual!) that makes a known pattern investable. Or else about being earlier to identify emerging, short lived patterns so that you can profit from more of their lifecycle. Teams make long term investments in technologies and strategies that help them do these things repeatably.
Post reply on HN