Hacker News Reader

Not the least because most of those data sources require a license.

Also they have huge amount of economic data, company financials, all somewhat normalised. That’s costly to maintain as I presume it comes from hundreds of different sources. Because these things break all the time. Broken feeds with wrong prices get sent, which require to reload the history when it gets corrected. A lot of google finance price feeds were broken and stayed broken for months They also integrate with trading desks, for instance who send bond pricing runs to their clients via bloomberg. The more banks you are client of, the more reliable your prices in bbg.

There also have quasi proprietary data, like the key terms of pretty much every bond issued in the world, that the issuers or book runners typically double check when they issue a new one. They create generic time series of on the run bond benchmarks, n-th future (which switched automatically when the future rolls), etc.

I am all for competition but agree, steep hill.

I'll just pile on here. The functionality of the Terminal is seemingly infinite and goes way beyond market data. Although it is obviously very good at market data. You pull up a bond, or future, or stock, or index, type FLDS and see an amazing amount of data beyond just the price.

It has maps (ships, refineries, fires, earthquakes, etc), flight schedules, overviews, top/bottom lists, correlations, regressions... ad nauseam. And also a quant platform, where you can code in Python I think, but I've never used it. I've been using for longer than I can remember and it's still impressive.

Also, they have generally a good to excellent help desk. Hit F1 twice and you're chatting to a real person, who can either help you or direct you to someone else who can. They have let me down on occasion too, but compared to other help desks/support, BB is the best.

I am starting to see more latency in pulling up functions in the Terminal, so I hope they're not losing sight of the UX. Near-instance response used to be commonplace, now I find that even some very common screens like WEI take noticeable time to load.

It is expensive. Many competitors have come and gone, and the closest thing now is I think Eikon, which also have installed to access the old Datastream data, and it's a complex monstrosity. FactSet is another one but I haven't used it in a long time. Every time cost-cutting time comes, one of the first questions is, "Do you really need Bloomberg?" Well, yes, I do.

Also don't forget about how deeply integrated BB is at many companies. Not just spreadsheets, but entire applications using the API.

So good luck, but I don't see it happening.

> Also, they have generally a good to excellent help desk.

I found they were good for ordinary "where do I find this function" type queries, useless for anything to do with APIs ("how does this call work") or data questions ("did you backdate this dataset").

The helpdesk explicitly doesn't support API questions, only for the Excel functions. I phrase my API questions to match in Excel BDP terms, that usually works. Data questions can indeed be difficult to get answered.

I have had pretty good experience with their API support. Generally, if we include our sales rep they will hunt down the right team to get our questions answered if we get the runaround the first time around. The places I've been at do spend a ton of money on terminal + other data licenses though.

"amazing amount of data beyond"

COKPMON Index is probably my fave in that department, you never know when you'll need to correlate something to the monthly kidnapping rate in Colombia

[Edit to update - Sorry, the link quote is/was from a different nick <cmacleod4> not yourself. My bad]

Just wanted to say - I loved a previous comment that included this -

"... Even if you could spend the billions to duplicate the infrastructure and rewrite all the software required, to actually do the same thing you would also need to replicate all the business relationships with data sources.

Disclaimer: I am not a lawyer/cat/financial-advisor/etc, but I used to be a developer at Bloomberg :-) " [1]

[1] https://news.ycombinator.com/item?id=26261167

I like the competition but I believe platforms should have a separation of concerns. Either provide a platform for data or a platform for forecasting. Bloomberg in my experience is mostly data and chat. The added benefit is that everyone has the same reference data. So if someone is trading agency bonds, they can reference a ticker and a yield and get to the same price. The yield itself may be meaningless (i.e. bonds with the same face characteristics could trade at very different yields), but the traders know the nuances and use the yield number as a round proxy and a way agree on price. Mixing both would likely result in having a half-baked product on both sides:

> Elsewhere, OpenBB can leverage deep learning to predict stock price movement using historical data, though in reality the model can be applied to just about anything, including economic data, crypto, and more. The company plans to double down on these predictive smarts.

This is dangerous. You shouldn't lead people to believe they can just plug in some symbols and automatically run a meaningful forecasting model. It may be true with natural phenomenon (e.g. when will be the first full moon 100 years from now), but not with complex financial data. You're leading naive users astray into thinking that some naive regression or ML model can give them insights when you're almost certainly falling into a trap (e.g. overfitting, using predictors that you don't have access to at the time, execution complexity, etc).

Create a service that's a better chat app. Another one to get the data. Another one to run analysis. And allow them to be interoperable through clear APIs. Don't try to build a magic black box that does everything for you.

> allow them to be interoperable through clear APIs. Don't try to build a magic black box that does everything for you.

One word ... licensing.

Its one thing to talk about data feeds for display on a terminal (and perhaps with an Excel plugin and whatnot).

Its another thing to talk about data feeds for "open" use by anything under the sun via API.

Multiply this by hundreds or thousands of data vendors worldwide who provide feeds to your Bloomberg-esque organisation and you see the problem.

Not quite dating, but they used to have GRAY, which directed you to a Sugar Daddy service.

what is GRAY? i tried searching for a source on this, but it wasn't terribly helpful as a search keyword

Bloomberg Apps are basically keywords that you enter in the terminals “command line”.

I suspect GRAY (because it’s capitalized…the BB command line is all caps) is one of those apps.

The more traditional way to represent that would be GRAY<GO>, which basically means typing G, R, A, Y and <Enter>.

They also have the tools for trading various instruments, handling RFIs etc.

I think people will be willing for this start up to succeed though because:

1) Bloomberg has an incredibly inept internal technology team. Dealing with them is a like a clown show. 2) Horribly aggressively sales-driven in all customer interactions 3) Everybody in banking management that signs the hundreds of millions of dollars per year invoice for Bloomberg absolutely hates them.

Current Bloomberg Pro user, same exact sentiment. Bloomberg just happens to have a preposterous number of barriers of entry for any aspiring competitor:

1) Licensing - historical data and metadata (need the day counting convention for random bond in a random country 20 years ago?) is heavily encumbered with licensing contracts. You're not allowed to take data off the terminal, for example.

2) Exclusive data access for things like macro sentiment numbers - for example banks and research companies submit their estimates for things like non-farm payrolls (earlier today)

3) Network effect for IB - huge amount of OTC trading is done via bloomberg group chats

4) Enormous library of pricing models and random practitioner-time-savers * along with the licensed historical data to be able to use them out of the box.

* eg what's the cheapest the deliver for a random futures contract you're looking at

The GIP GP and whatever other price plot function that the OpenBB project is emulating are barely worth mentioning in this context.

The unified data is great. The basic and advanced charting is very good. The basic analysis stuff is very good. Complex pricers end up get built by quants, pulling in the bloomberg data - on 100k++.

300,000 users globally, a high percentage probably use little more than a basic chart, news, chat, MVP and POSH.

> 300,000 users globally

Interesting.. if we're assuming these are customers all on the same license, that would put the revenue at around $6B per year for this.

I wonder if there are any other tools within an order of magnitude of that user count that generate that revenue.

I think the way to compete here is not directly targeting BT customers, but looking at other potential customers that might not be able to afford a BT, at least in the beginning of their journeys, and use that revenue to fund "completeness".

Also important re chat that most firms want a /very/ short list of approved apps to manage for inter firm chat. Typically it would look something like Bloomberg, Refinitiv Eikon, and Symphony.

Symphony has to be one of the worst chat clients I have ever used.

You may not have had the pleasure of using Skype’s ‘persistent chats’.

Reminds me of Emacs being an entire OS

IIRC, the bloomberg terminal did have march madness and world cup pools (well, for bragging rights, not for money) built in :)

> And then they have an approved chat function

Curious, who do you chat with? Is it an intra-company chat? Or wider than that?

It's instant chat between every terminal subscriber. It's very popular.

Let's say I want to trade a large amount of $TICKER.

Sometimes, just sending a buy order is not ideal as it might drive prices up.

So you start calling brookers and hope they help you by finding other folks who want to sell large amounts of $TICKER. They also might spread orders over a larger time frame.

That's just an example, but overall the chat is used to negotiate assets. Buying and selling stuff is often a very manually coordinated process.

The chat feature is hugely important for traders. I've seen traders use it to iron out the details of OTC deals prior to booking the trade. It was used so heavily for this purpose that I know one bank even used the data as part of their MiFID II regulatory reporting.

As others have mentioned, the chat client (Instant Bloomberg, or IB) is used for a significant amount of financial trading and information exchange. It went down for a couple hours at one point which had ramifications on financial markets.

https://www.businessinsider.com/the-bloomberg-feature-every-...

It’s kind of funny, but IB (the chat function) was basically slack 2 decades before slack ever existed.

It’s a chat function that can connect any Bloomberg terminal user, and you can also run “bots” on.