Show newer

True, and also: how is this not an abuse by Twitter of its dominant position @randypicker?
RT @joshgans: How is blocking Substack in anyone’s interests? Twitter included.

🐦🔗: n.respublicae.eu/lugaricano/st

Is Musk conducting an experiment to measure the strength of Twitter's network effects?

One more crazy decision: you cannot like or RT substack links any more.
RT @paulg: You won't be able to like or retweet the second tweet in this thread, which links to an interesting essay by Scott Alexander. You can however like and retweet this one!

🐦🔗: n.respublicae.eu/lugaricano/st

RT @alexandrosM: STOP THE PRESS (like, literally)

A simple demonstration of how GPT can change the news, forever.

Prompt: "What can be changed in the article below to make it more informative, truth-focused, and neutral?"

Left: original text
Right: rewritten text

Read🧵 🔽 for more thoughts.

🐦🔗: n.respublicae.eu/lugaricano/st

In my view, the bottleneck is not the writing, but the reader's limited time. Thanks to AI, there will be more writing, but also more plausible fake news. Good journalists will be even more important to ensure veracity and importance of the article.
RT @greg_ip: To paraphrase an old saying, technological disruption is when your neighbor is automated out of a job; the robot apocalypse is when you're automated out of a job. Could ChatGPT throw journalists & other pr…

🐦🔗: n.respublicae.eu/lugaricano/st

RT @sbuhai: @lugaricano Ja, ze 🇩🇪 coast guard eventually grasping that distress call...
Feeling shameful is in good order but not compensating: many deaths/ other horrors could've been prevented had it moved earlier, both concerning sanctions against 🇷🇺 and sending weapons to 🇺🇦.

🐦🔗: n.respublicae.eu/lugaricano/st

Well done, Vice- Chancellor Habeck.
RT @Telegraph: 🇩🇪 Germany’s “deeply ashamed” vice-chancellor has told Volodymyr Zelensky he is sorry that Berlin took so long to send weapons to Ukraine.

Read more here ⬇️
telegraph.co.uk/world-news/202

🐦🔗: n.respublicae.eu/lugaricano/st

RT @whyvert: New paper: before the Counter-Reformation, Catholic and Protestant cities had comparable numbers of scientists per capita.
Afterwards, Catholic cities experienced a persistent relative decline.
Counter-Reformation's search for heresy was a negative shock to science.

🐦🔗: n.respublicae.eu/lugaricano/st

India: Establishments with high demand for AI skills show negative effects on postings and wages for:
-High skill, professional occupations
-Non routine work
-Analytical/communication tasks.
By @alexlcopestake, Marczinek, @ashley_pople,
@KathAStapleton
steg.cepr.org/publications/ai-

🐦🔗: n.respublicae.eu/lugaricano/st

@HansbergRossi @klausvanieper @jlcruzalv @RevEconStudies Congratulations! It is a fabulous paper.

🐦🔗: n.respublicae.eu/lugaricano/st

@zingales I do think it is a huge deal. If you had to mark to market all bank books (which would be the inevitable consequence of a continuous "bank walk", unless as you say in paper a deep recession comes to stop IR increases) you would be needing capital throughout the banking system

🐦🔗: n.respublicae.eu/lugaricano/st

RT @R_Thaler: Interesting thread and blog post. Theme: mobile apps make deposits less sticky. Untraditional but helpful take on “transaction costs”. Need to include clicks, time, and other forms of sludge.

🐦🔗: n.respublicae.eu/lugaricano/st

Huge implications in my view (not authors'!):
- Not a run: not about deposit insurance.
- No reason for HTM exception. Banks must mark-to-market their assets.
- Fed faces new challenges in controlling inflation without crashing the economy.
Link: promarket.org/2023/04/04/desta

🐦🔗: n.respublicae.eu/lugaricano/st

Show thread

In the companion paper, the authors find that a 400bps increase in the fed funds rate, roughly what the Fed increased rates in 22, leads to a 9.6% drop differential relative to trend in deposit growth for non-digital banks without brokerage fees and 15.2% with brokerage fees.

🐦🔗: n.respublicae.eu/lugaricano/st

Show thread

Banks with neither an effective mobile app, nor a brokerage account see their deposits grow at a steady 1.5% rate in the third and fourth quarter of 2022. By contrast, banks with both a functioning mobile app and a brokerage account see their deposits drop by 2% in Q4 2022.

🐦🔗: n.respublicae.eu/lugaricano/st

Show thread

Where is the money going?
Figure 2 reproduces Figure 1 but now split into four groups: Digital and non-digital and banks with brokerage fees versus those that don’t report these fees.

🐦🔗: n.respublicae.eu/lugaricano/st

Show thread

From April 2022, when the Fed started raising rates, $860 billion of deposits left the banking sector, mostly into money market funds. Two-thirds of this happened before SVB collapsed. This is a silent walk that can’t be stopped by deposit insurance.

🐦🔗: n.respublicae.eu/lugaricano/st

Show thread

More than half of the banks have introduced a mobile app. Figure 1 shows shows the quarterly average growth in core deposits, computed separately for digital and non-digital banks. Digital banks start experiencing deposit outflows in the third quarter of 2022.

🐦🔗: n.respublicae.eu/lugaricano/st

Show thread

Bank regulators assume that deposits are sticky, meaning they don’t move around much. But digital banking is changing that. New research, described in a blog (link below) by Koont, Santos, Zingales shows how mobile apps make deposits less stable.

🐦🔗: n.respublicae.eu/lugaricano/st

Show older
Mastodon

A Mastodon forum for the discussion of European Union matters. Not run by the EU. Powered by PleromaBot, Nitter and PrivacyDev.net.