Basically the whole market looks overvalued as shit, and half of you are going to short it and lose everything, and half of you are going to long it and lose like 20%. If dogshit industrials and deep value and utilities start pumping, or god forbid EM trash is pumping, it means we're probably closer to a top than most want to admit, but who the fuck knows.
Too many dollars chasing too few things.

I think few people here talk about their losing trades and the crummier side of prop trading.
I'm not here to sell you anything. Frankly, I would like to convince more of you not to pursue trading, because it's one of the worst work-life balances you can imagine. Let's talk about some losing trades I've done recently.
I am not going to name tickers because it's irrelevant anyway. 1) The biggest loser thus far, recently, was a pseudo-arbitrage between a well known commodity and an infamous holder of the same commodity. I don't really need to share which names; you can guess.

The thing about mean reversion trades like this is they can be very, extremely painful. Trend is nice and gentle. Mean reversion you need to be prepared to be ripped open and then bet more with your own blood.
Risk management tends to be far more important on these trades.

I'm not saying crypto looks like a local top here, but I am saying that in a month if it's 20% lower it will have been entirely obvious to sell here.
---
That said, I just follow momentum, so yeah.

https://twitter.com/nope_its_lily/status/1732809708648734874

My own signals and positioning is quite long, but am I uncomfortable here? Absolutely yes.

Good post-mortem from in my view a talented fundamentals trader
Sizing is generally what kills you, and I've never been able to find a sizing on bets as robust and good as simply 1/n

https://twitter.com/kingfishcap/status/1764195121875362234

I saw a few of his write ups and I can't disagree with his line of thinking, but two things:
(1) It's hard to critically own value or even special sits stocks in size especially without positive momentum simply because you're giving a giant F-U to the market. The price of any asset is the collective intelligence of everyone else, so buying something in a downward trend is simply saying your thesis, your view, is not only better than everyone else's, but that you know eventually it will be realized
(2) Selling options is a mine-field, and whatever your worst case is, you should at least double it. Selling puts of course is less risky than selling calls.

One interesting thing I talked today about was viewing risk, especially in the context of trading relatively sketchy (be it shitcoins with rug risk, foreign equities on 3rd-tier exchanges, etc.) assets.
Risk is three-dimensional. It is:
- Exposure time
- Concentration
- Loss potential (probability of risk event * recovery value)
In general, it can be +EV to take relatively high credit risk for short time horizons, especially if the expected return makes it worth it and your concentration is fairly low In general, I prefer always to take more risk with a smaller amount (whether it's junk bonds, Chinese equities, or F-tier shitcoins) than take less risk with a larger amount (aka like BB bonds or ADRs) because risk is highly correlated in black swan type events, and assuming you're long the asset, your loss potential is a lot lower in the extreme left tail

This is an interesting question and kind of goes into the divide between a discretionary approach and a quant one.
In reality, factors aren't real things, and over long periods of time, a company or even more, an index, can shift its actual factor exposures.
If I make a call like "long e-casinos, short casinos", I'm not taking an outlook on any given factor, I'm taking an outlook on a business trend.

If I were to quant it, I'd compare it to two benchmark portfolios - a risk parity one (equal expected vol contribution from all strategies) and an equal weight one (could be equal net or equal gross, though, since sometimes you're running long/short ideas). I work with a very talented quant here and we've discussed this exact problem - in a multi-factor risk model, tweaking it based on a PM's outlook.
There's no right answer here. For me, currently, I just equal weight roughly all strategies, constrained to a $ gross cap and $ gross min (essentially 0). So the trend following layer occurs in between the min gross cap and max gross cap per strategy.

More updated portfolio breakdown:
- Long high coupon preferred and qualified dividend stocks, skew towards consumer staples and energy (oil)
- Long new GLP1s, short old GLP1s
- Long medical devices, short fast food
- Long China rebound (calls + tech stocks)
- Long Taiwan semis + American tech, short Indian offshorers
- Long hotel chains, short hotel owners
- Long airports, short airlines
- Long Pakistan (not positioned yet, but closer to election date will be)
- Long Phillippines (stock transaction tax)
- Long bitcoin (crypto, tactical)
- Long Japanese net-nets with good momentum (1/3rd infotech, 1/3rd industrials, 1/6th fin, 1/6th consumer disc)
- Short vol discretionarily (mostly single stock, haven't ramped much here)
-----
Probably a lot of factor mismatches here, but most of these trades are relatively idiosyncratic, and I'm not particularly concerned about being paid for factor risk here.
More worried about being overdiversified.
Overall portfolio is skewed short dollar, long beta (but not much). I'm going to spend a lot of time tracking the performance sleeve by sleeve of these different verticals. There's some sector skews here on the sleeve level of course (+med devices/-fast food hs residual XLK/XLP beta), but that's again not something I want to remove.

I find it helpful, mostly because I'm a miser and pessimist at heart, to think about events in terms of short term, medium term, and long term flows impact.
For the Bitcoin ETF launch (basically a guarantee):

** Short term **:
I expect bullish flow, if any, from speculators primarily, similar to the launch of BITO in 2021. However, this is short lived and exponentially decaying.
I do think this will mark a topping period similar to the late April and October 2021 periods, where Bitcoin related beta bets tend to outperform in the short term, as well as altcoins.
I do not have any compelling reason to expect significant, durable flow from otherwise non-crypto participants in the short term; there is no catalyst that would imply it.

I do expect varying amounts of bearish flow from speculators unwinding their bets. We can see, despite the narrative that "the consensus is sell after ETF", this isn't actually reflected in any meaningful market positioning. Funding, volatility skew, and CME basis all expect high upside and high volatility.

There is, potentially, a catalyst with GBTC conversion, which may lead to redemption (unclear). ** Medium term **
This hinges primarily on if the spot ETF is meaningful, and moreover, if the spot ETF is followed by spot ETF filings for other cryptocurrencies.

I do think the medium term flow post-ETF will be from more institutional legitimacy of crypto, and more investor interest. I do not think that increased investor access will be meaningful here (as I asked about before).
That said, I think the Blackrock and Fidelity names will be large here for increasing both investor awareness of cryptocurrency, as well as driving more financial advisors and retail end users to consider it to be part of portfolio allocation (however, this will take a *while*).

I think the more meaningful case to consider here is the impact of further spot crypto ETF approvals conditional on substantial trading volume for the bitcoin spot ETFs. I do expect that one or two spot ETFs will absorb most of the volume, and it may not be the cheapest (see GLD vs IAUM).

I think that this is generally bearish for traditional crypto exchanges, since the custody business is much worse than the exchange business. Others may disagree, and the bullish outlook is that this may serve as a loss leader to drive altcoin interest.

That said, my general expectation is the medium term flow impact would be neutral to bullish, with more likelihood of slightly bullish.

You've finally become a real adult when you realize that no one's controlling the strings.
There's no shadowy cabal.
There's no faceless Illuminati that somehow orchestrates the world.
It's just people. Mostly incompetent people. Uncoordinated groups of us, meandering and trying to maximize our own happiness and comfort.
There's no one in control, at all. And that's scarier. It's much nicer to imagine, even if you conspiracize they're the ultimate evil, that somehow, Davos or WEF or UN or the Jews or Big Finance ETFs are in charge.
Then you can believe the world makes sense, that there's ways to restructure society that can actually make it simpler and better.
Because without that, your efforts seem small.
If there's no one behind the curtain, what are you fighting for exactly? What can populists rally against?

Some thoughts from a relative neophyte in crypto.
I am sort of adjacent to the space, but not in it exactly...
I remember I first heard about crypto in 2017 and opened a Coinbase, bought some ETH and BTC then, sold right around the top for a small loss (bought around the top too).
I do remember the 2020-2021 cycle very well though. I made a bit from it, but not much, personally (it's just not a big chunk of my assets).
Crypto is a very rhythmic asset class, and one of the best to trade. @cobie explained it well a while ago, but it's dominated by narratives and momentum of narratives.
Right now arguably the narrative is the ETF, but let's hope it continues to sustain, because this exact same rally has failed twice before (2018 - top of the market/introduction of CME futures, Oct 2021 double top - launch of Bitcoin Futures ETF) 1) There's actual signs of institutional interest - both from the grapevine and also quite notably in the leadership of CME as the venue of exchange for Bitcoin trading. That's a fairly new thing, and good for the space.
There seems to be renewed interest in Bitcoin as a store of value or portfolio diversification trade, and so far notably this year it has proven some merit (Mar 2023 - the easiest do-or-die moment for Bitcoin).
It's unlikely the actual launch of the Bitcoin spot ETF will bring much inflow into the space, but I could be wrong. Gold had similar in 2004, but it really depends on a relatively unscrupulous network of financial advisors to push said ETF onto their clients (gold was an easier sell, because well, everyone knows what gold is).

One carry trade which is fun sometimes is buying long dated straddles and shorting short dated straddles. This tends to be cleaner when the underlying has no real drift (so all moves generally mean revert), otherwise you have to regularly restrike it (this is the benefit of var swaps, comparably). btw to clarify this trade:
(1) this really only is carry when the ivol curve is in backwardation obviously, otherwise it's negative carry, right
(2) this pays you in two potential ways:
- long dated vols blow out (good)
- the shape of the volatility curve doesn't change much over time (the carry aspect)

A thread on alphas, strategy, and proprietary trading.
Half of my cool thread ideas come from my talks with @cam_perrault, mostly chastising her for taking too much risk.
@choffstein has a good saying on this, that risk cannot be destroyed, only transformed.
And why is this important for thinking about alphas. A strategy, no matter what your approach is, can be understood as a combination of alphas of varying quality/strength.
We have to decompose here what we mean by alpha.
The best way I've heard it summed up is by Paleologo - an alpha can be timing (your ability to time a specific instrument), selection (your ability to identify a particular instrument), or sizing (your ability to size bets among alphas, kind of a meta-alpha).

If you're a profitable trader on your own capital, you need a certain amount to make it reasonably worth it compared to trading someone else's capital.
If you trade a 50mm book for someone and take 17% payout split, let's say 14% realized after fees and other costs, if you return 30% on that, you're taking home $2.1MM.
If you did the same on your own capital, you'd need (assuming again generously no fees and the same 30%) $7,000,000.
And of course, when you trade OPM, you're not risking your own capital.
And you usually have a salary of some sort to fall back on.
It's very difficult to make it trading your own money. There's certainly opportunities (on-chain in crypto, prediction markets) where scale invariance isn't a thing, and there's opportunities to make good money in the thousands of dollars where it makes no sense for a bigger firm to trade it.
But it's fundamentally difficult once you compute the actual risk (volatility) and time you spend to really make trading your own capital worth it under a few million dollars.

don't do it the way I did, and try to find some good firm to clerk for to start with, or preferably, go be a software engineer/data sci and make risk/stress-adjusted more money

https://twitter.com/astropineapple1/status/1700994691939106849

I'm still barely good at things, and realistically most people are going to fail at trading because trading is a zero-sum game.
I've talked to multiple people on this and if you're not dissuaded, the best way to go is to try and be mentored somewhere great.
There's a few true bandits here (like @cam_perrault @ darrin (where is he?)) who have made it without institutional experience, but it's very hard and the worst part is not learning good habits early on, which will only compound/be harder to break later on

How to analyze a strategy:
I've had this chat now with two different and dear people, relating to opposite timeframe strategies (one very long term, the other intraday).
How do you figure out if your strategy is:
- Real
- Good The classic metric of strategy measurement is Sharpe, which is a fairly strong measure of risk adjusted return, return per unit volatility. It makes intuitive sense -- volatility is a measure of risk, and other than true arbitrages (rare), any return requires some level of risk.

A thread about macro quantitative techniques.
Systematic macro is kind of important even for passive portfolios for three reasons:
1) Portfolio performance is regime dependent
2) Portfolio performance is factor dependent
3) The real economy matters on long timescales **Regimes**
Regimes aren't real things actually, but you can sum them up as "big things that tend to matter a lot". The classic example is interest rates and the broader yield curve, inflation, etc.
The issue with regimes is you can overfit to combinations of macro factors.

I wanna eventually convert this into a brain dump blog post but I wanted to write something about broken business models, at least to share where I have failed or where I’ve reflected on others who have.
So this will be a list I’ll try to add to time to time. #1 - The Success Paradox
With social networks and App Store like models, a popular business format seems to be “X on top of Y [some closed eco]”. A classic example is Shopify integrations. Good business right?
No. The issue with this model is you are or will be a victim of your

A really interesting take I heard recently about last year's market movements was that the equity/rates comovement was a story of rebalancing and convexity.
A change in yields from 0.25% to 1% is more substantial for bond prices and reallocation than a change from 4% to 5%. In the first part of the year, as bond prices fell significantly, bonds as part of target date and other risk parity-type strategies shrank in terms of the overall portfolio value, leading to equity selling.
This is also consistent with market inflection points, including Jan 22.

Going to share some time today a synopsis of May's performance of the Lily Forever Portfolio. May was a mixed month. LFP performed okay on an absolute basis - at current daily performance, looks to be around -35bps MTD vs 30 bps MTD for SPY alone. We'll see by close time. The biggest detractors have been the TLT allocation -- a better designed variant here would reduce the negative convexity in TLT (perhaps replacing it with a differently weighted AGG, skewed to higher duration) and XLP, which has had a terrible May.

In the most important update of your day, Lily’s Forever Portfolio again pulls ahead year to date of SPY, 8.4% to 8.3%. It realizes a 2.36 Sharpe ratio YTD to SPY’s 1.5%. I will try soon ish some dynamic rebalancing stuff and also replace the gold weight with a more comprehensive commodities breakdown and see if it changes positively the analysis

One critical thing to understand in tactical allocation is rebalance timing including how to initially allocate.
Important point:
Since May 9 2020 to now LFP has trailed SPX by 16%-ish
Since Jan 1 2020 LFP has beat SPX by around 2% This is because by design the LFP is defensive and long term outperformance comes from low downside capture and decent but not full upside capture. So if you were very unlucky and allocated on May 2020 you’d hate me!