1 like = 1 finance/markets/trading concept described in a tweet
Also, am taking requests
Also, am taking requests
Markets are made up of people, with different agendas. Big, small, aggressive, patient. They reveal their intentions by how they trade -- patterns in order placement and execution. A healthy market has a large diversity of animals in the zoo.
Cc @pauld44
Cc @pauld44
HFT, really any short term trading (under a few days), is about predicting temporary imbalances in supply and demand.
e.g. Lots of demand just showed up. If you can predict supply is coming soon, you'll want to bridge the gap for a small profit.
e.g. Lots of demand just showed up. If you can predict supply is coming soon, you'll want to bridge the gap for a small profit.
Ants: make up the majority of the activity, moving small amounts around at a time. No bias.
Mules: they need to keep buying every day, it's in their rule book.
Vultures: looking for situations where everyone else has given up.
Mules: they need to keep buying every day, it's in their rule book.
Vultures: looking for situations where everyone else has given up.
Don't hear of much success applying big neural nets to market prediction. It can help on the margin, but not as a major alpha source afaict.
Other aspects of "ML", which we like to call *statistics*, is bread and butter in quant trading.
Other aspects of "ML", which we like to call *statistics*, is bread and butter in quant trading.
Stock prices are weird. The "last price" is the price of the last trade. The real state of reality is a stack of orders willing to buy and sell at various prices. Occasionally they match with each other and emit another trade.
Did I misunderstand q?
Did I misunderstand q?
Many quants are sampling some techniques from ML. Most of ML, in its current form, is ill-suited for time series prediction. There is probably some stat innovation unique to trading.
🧱 Options. Just like in everyday life, an option is... The option to do something. There are lots of types of options. The important thing to know is options are valuable, but only under specific conditions.
You don't care about the option to go sailing if there's a storm.
You don't care about the option to go sailing if there's a storm.
Yes! I'm not expert but many hedge funds derive practically all of their alpha from alt data.
Random idea for alt data: Google ads being shown for a private browser for specific keywords
Random idea for alt data: Google ads being shown for a private browser for specific keywords
🧱 Futures. A promise to purchase, or sell (i.e. To deliver), a particular thing at a particular time in the future.
If you are farming corn, and you plan on having 10k bushels in May, you can lock in the price that you will get today by selling a May Future.
If you are farming corn, and you plan on having 10k bushels in May, you can lock in the price that you will get today by selling a May Future.
Beats me. Can't get much faster, and many are dabbling in custom chips already. I always wondered why exchanges didn't lease compute cycles directly on the matching engine. Maybe that's the future.
Cc @dendisuhubdy
Cc @dendisuhubdy
I mostly like the old school investment guys. People in trading with good ideas don't tend to be very public about it, so hard to say. Dalio seems pretty out of it, but they weren't really a trading firm. Simons is obviously still a God.
Yes. Location doesn't really matter if you can attract the talent.
If it's truly risk free, then mostly just how non-obvious it is. Otherwise, stuff like access to capital, especially leverage. Also, how many competing opportunities that are better.
Yea, TA is just patterns. But humans use heuristics and patterns for decision making, even if they don't believe in TA. Rounds numbers are magical to people. Anchoring effects, etc. There are more direct ways of finding this stuff than lines on charts.
I'm not sure. I would look outside of crypto, towards less liquid assets. Maybe smaller countries, or personal loans (eg LendingClub has an exchange, I believe)
Yea most of what I see is startup-esque. Probably just following the availability of VC cash
🧱 Beta is the returns that everyone has access to. Like investing in S&P500. It's the"commodity" of investing. It is available cheaply and most people's portfolios are mostly earning Beta returns. Beta is second to alpha...
🧱 Alpha is the returns that are above and beyond the readily available Beta. It's not correlated with beta, and gives you more return without as much additional risk. It's the holy grail of investing, and the concept can extend to business, in general.
🧱 I've seen all kinds of algo trading. Everything from simple TA rules to self-adapting prediction machines using big data.
In general, it's data + rules in, and trades out. No human in the way, except for designing and building.
In general, it's data + rules in, and trades out. No human in the way, except for designing and building.
Yes, microwave networks are most definitely a thing. Huge competitive advantage for the firms with the fastest links between key venues.
🧱 Options skew is basically the market predicting how future expected volatility will change conditional on prices changing. With stocks, if prices fall, they will tend to become more volatile. The options market prices this via IV skew.
🧱 Quantitative easing! Central Banks control short term interest rates. When those go to 0, they start controlling longer term rates. This involves buying stuff like treasury bonds, taking them away from banks and giving them Fed dollar deposits to lend
🧱 Repos! Simple concept, fancy term. Many investors and traders use leverage, lots of it. One way of leveraging is via overnight repos. You put up an asset as collateral, typically something very safe, and can borrow overnight against XX% of it
That's a tricky one. I don't have strong feelings. Probably I'd want to regulate leverage -- reduce it for non-bank entities, especially. Also, maybe intratrading in general -- but I don't know how to curtail that w/o weird side effects. Shorting is good.
🧱 Debt. When you need cash, sometimes you can find someone to bet on you -- they're willing to give you money now, in exchange for a bit more money later. They take the risk of you not having the money later. You get money you need now.
🧱 Almost all derivatives will have some embedded leverage. Typically, you take an implicit exposure that's larger than the amount of cash you put up as collateral. Derivatives can be scary because of how quickly they wipe out that cash collateral.
🧱 Banks take in deposits and give out loans. What happens if those loans go bad? They need a capital cushion to take losses and protect depositors. This is equity (stock). Their capital ratio is basically the ratio of equity to loans.
🧱 Liquidity -- you can have something really valuable that you just can't sell right now. Your house is worth something, but is completely illiquid in many situations. What's it "worth"? If you need cash in the next 5 minutes... maybe nothing?
🧱 Solvency -- if there are debt payments coming due and you won't have the cash to pay them, you're in trouble. Market prices and liquidity matter less here, unless your cash is going to come from selling assets. In which case, illiquidity → insolvency!
🧱 The "Money Market" is the market for short-term "commercial paper", which are generally very low risk short-term loans taken out by big companies. This is like GE borrowing $100 mil for 2 weeks to pay payroll. Why? Holding cash is inefficient.
🧱 Rule Number 1: avoid losing all the money. Avoiding risk of ruin is the most important thing you can do.
Don't take my word for it, here's Stephen Shwarzman:
Don't take my word for it, here's Stephen Shwarzman:
If you have lots of alpha, basically none. Spin up an IB account and trade from there.
At the other extreme, if you're competing mostly on speed, then FPGA, or maybe some custom hardware. Million bucks per year?
At the other extreme, if you're competing mostly on speed, then FPGA, or maybe some custom hardware. Million bucks per year?
They could provide better strategies than what they do now (see: @breakingthemark ) but I don't think it's a great business. Typical people don't pay for this, they pay for babysitting (telling them not to sell)... so they'll become plumbing, maybe?
There's no right/wrong answer. Depends on circumstances. Sometimes you want to consolidate and create optionality, other times it's better to protect your downside and put all chips on making a single outcome happen.
🧱 Money printer go brrrr. The Fed is buying assets, some of dubious quality, and putting the cash on Bank balance sheets. IFF the Banks start lending this cash to business/people, then more money pumps through the economy. Eventually, you get inflation.
Haha a question for @MutinyFund. I don't see how typical bubble dynamics can play out with index funds. But I do bet there is more return available to active management than before.
Probably? It's a good Q. In the wild west of crypto, sheer institutional momentum might be enough to keep T liquid even if it turns out it's not backed by anything. As if any of these shitcoins are backed by anything, anyway
🧱 What is money anyway? 💵 I think it's pretty simple. Money, or wealth, is a symbol of another person's promise to do something for you. That doesn't mean they gave it willingly, or that you earned it. Nevertheless, if you have money, someone feels like they owe you something.
🧱 Spread trading. A billion factors impact the S&P 500. A fraction of that drives the difference between, say, 2 similar oil company stocks. It's often easier to find alpha predicting that more specific thing, and betting on the difference between two stocks. 
🧱 Stock buybacks. A company can give money back to shareholders by buying back their shares. They will hold an auction, like a reverse IPO. This reduces the quantity of outstanding shares, which means higher $ value per share (higher stock price).
🧱 Market prices are backwards looking -- they are historical trades. The true state of the market is represented by its Bid/Ask. The best bid is the highest price someone is willing to pay at any given time, and best ask is the lowest price willing to be sold at that same time
Quant trading has mostly consolidated to a few big firms. It's a "late stage industry" and is probably in decline. Would be surprised if it had substantial new impacts going forward.
🧱 Covered Calls is an options strategy where you own the stock, and sell (short) Calls on top.
Owning stock is equivalent to +1 (long) Call and -1 (short) Put. Short a Call on that and you are just left with being -1 Puts. You probably don't want that.
Owning stock is equivalent to +1 (long) Call and -1 (short) Put. Short a Call on that and you are just left with being -1 Puts. You probably don't want that.
🧱 The Fed[eral Reserve] controls interest rates by buying/selling stuff -- mostly treasury bonds -- with the goal of keeping inflation and unemployment "under control". They also lend to banks that are desperate for cash on a short-term basis.
🧱 Shorting a stock is equivalent to +1 Put and -1 Call. Buying a put is just +1 Put. So the difference is -1 Call. Shorting a call gives you some cash up front, but exposes you to unlimited losses if the price of the stock goes up.
Shorting means you don't need to put cash up front. That cash up front, to purchase the put, can be really expensive when markets are crazy. The option is valuable! You pay for that premium, and it erodes over the life of the option.
🧱 If the Chinese Yuan falls in value relative to USD, Chinese products become cheaper in USD terms -- they will accept less USD for the same thing. This means more demand from USD holders. This means more exports from China, which is good for an export-driven economy.
🧱 Central banks try to manipulate their currency. Sometimes to make it cheaper (by "printing money"), to increase exports. Sometimes to make it stronger (by selling other currencies they hold and buying their own). This is not inherently bad.
🧱 It's easy to make your currency cheaper. The risk is inflation for your citizens. It's hard to defend your currency -- eventually you run out of stuff to sell to buy it with, and the price of your currency collapses.
🧱 Sovereign defaults on local-currency debt can and does happen. Will they default on their own citizens, with mostly $ denominated debt? Probably Not. Will they default on China if SHTF? Sure could. next-finance.net/IMG/pdf/Why_So…
🧱 Binary options are options that expire at either 0 or 1 depending on whether a particular outcome is achieved, e.g. S&P500 closes above 3000 on January 2021.
Just like any other product, if you can predict it, you can trade it profitably.
Just like any other product, if you can predict it, you can trade it profitably.
Increased moral hazard, eventually inflation for USD. But it might take a really long time
Maybe not black swan... but Emerging Markets could have sweeping defaults which would lead to regional military action
Buybacks and dividends both give cash back to shareholders. But buybacks do so in a way that increases the share price. This is especially appealing to executives who has stock options. You can see the potential for conflict of interest. Probably overblown
Countries issue debt to fund all kinds of stuff, famously wars. If they're lucky, they can issue in their own currency. Sometimes, they'll do it in another country's currency, in return for a lower interest rate. This can be hard to pay back later.
🧱 Risk on/off. All assets can be boiled down to a single bet: either long volatility, or short it.
Risk on = business as usual, desire for incremental returns, let's short vol
Risk off = protect what you have, or get protection to disaster (+vol)
Risk on = business as usual, desire for incremental returns, let's short vol
Risk off = protect what you have, or get protection to disaster (+vol)
Yes, I know a few that make $500k-2mil range, but the club is small and you need to keep adapting to keep up
🧱 Volatility is how much the price of something moves around. Bitcoin is much more volatile than the S&P 500. More volatile assets tend to have more uncertainty associated with their future -- they might be worth way more, or way less.
🧱 Credit Default Swaps (CDS) are insurance contracts that pay you if a particular bond or loan defaults. Like life insurance, you pay X/year in premiums, and get Y payout in default.
This is a way of making a short bet on... anything that can default
This is a way of making a short bet on... anything that can default
🧱 Value investing is a strategy that looks for stocks that trade below liquidation value. The core idea: if you can strip it and sell it for parts for more than what it's worth now, it's probably a bargain.
🧱Once you're convinced something is going higher, you are eager to find evidence to support your thesis. Bull traps are buyers rushing in at the first sign of a move up, after a prolonged decline. About 50% it will be a head fake!
🧱 I view gambling as entertainment. If I don't have alpha, and I'm not investing in an asset that will generate future cashflows, I wouldn't view it as an investment. It's fine to do, but I'd keep it separate from an investment portfolio.
🧱 Options are typically struck at round numbers, and can lead to distortions in trading around those prices. I think the majority of "whole number" effects are due to psychology, moreso than options dynamics, but it's hard to say!
🧱 Corporate governance is the system of formal, and informal, policies and relationships that keep a company "on rails". In theory, companies should work for their shareholders. In practice, it's complicated. Everyone has their own incentives.
🧱 Hedge funds invest wherever, and however, they think is best. They will use leverage, and take risky bets. A good hedge fund will generate uncorrelated returns to the market. If they're trading stocks, that's often by hedging...
🧱 A hedge is a position taken to protect or counteract some aspect of another position, or an entire portfolio. For example, if you're long Pepsi and you want to hedge against general economic conditions, you can short the S&P 500 against it
🧱 Alpha is a generic concept, it applies outside of trading. I like Peter Thiel's formulation: everything good is built around some kind of secret that isn't known or fully adopted by competitors. As in investing, so it is in business.
Still probably HFT, followed by any hedge fund. Check out Two Sigma, Jump Trading, Tower Research, etc.
🧱 CFOs are like fancy accountants for big companies. Mostly they make sure the money is coming and going like it's supposed to. This requires increasingly more sophisticated tools and processes at scale.
Historically, yes, real estate has performed relatively well in inflationary environments. The one caveat I would add is that inflation will mean higher interest rates, and higher interest is not usually good for real estate prices. I prefer Gold!
A thread on buybacks vs dividends in the context of executive compensation and COVID bailouts 🤯
If you're ultra latency sensitive, it will be on-prem for a while. Otherwise, seems mostly legacy issues holding them back. Lower latency is a secular trend.
Most hedge fund AUM is chasing betas at this point. They will increasingly consolidate, commoditize, and be forced to compete on price (goodbye 2/20)
🧱 Debt deflation is when debt becomes more expensive to pay back over time. This happens when the money itself becomes more valuable in "real terms" → the same $1 buys more bread.
So the $100 you owe tomorrow is worth more to you than $100 today.
So the $100 you owe tomorrow is worth more to you than $100 today.
I think it's still a good place to start and build fundamental skills. I would move to a tech company with a product after 2 years, though
