Q&A: Options Trader at Top Prop Firm

I’m probably the wrong person to ask, I don’t dabble really in my personal brokerage accounts.

And I don’t want to be that lame ass saying to just put it in an index fund or a diversified portfolio, (because that is the long run answer).

Maybe I’d pick stocks I believe in, in terms of how I view the future. Like Warren Buffet bought coca cola way back when, because it tasted good.. so something similar. Then you get the ride the wave of something that excites you.

25k just isn’t enough for it to be anything than a fun game..

But me personally, I think I would start an internet business, e-commerce or something instead. Globally, 3 billion people don’t have access to internet, but soon will.. startup costs for e-commerce to be successful is super low and if you did the math “trader talk” says this “strategy” has a super high Sharpe ratio..

It is true that big firms gobbled the small firms because of scale and technology and capital. But honestly, I think that is over now. There are actually too few market participants. You see this based on trade allocation. Too few people take on too much risk once something trades, you see them trying to reverse out of it all the time. This can actually be dangerous because the market needs liquidity diversification. The people I met at the startup worked for the guys who started Long Term Capital Management, a historical hedge fund, the story is fascinating. And I was being hired with the notion of becoming partner. There’s nothing more exciting than watching something grow from small to big and being part of the reason why with people you admire.. it was a no brainer to switch. (I talk about this more in the podcast).

This is the book about the Hedge Fund Long Term Capital Management

https://amzn.to/2z1DVcl

If you can talk the specifics on their successes and failures, you'll have a really easy way of sounding smart when talking about macro finance and trading.. it helped a lot when just talking back and forth in interviews..

Tricky question..

Can I ask, why do you want to switch? What specifically interests you about Prop Trading? How old are you?

Prop firms generally only hire people with specific trading backgrounds that fills a need for them or someone who is young at the entry level stage who can still be trained. Most top firms want to get kids out of college who join a rotational training program as a trading assistant. That being said, my route was the most non-traditional, so it’s still not out of the question.

First, I would highlight your math/stats background and any technical skills you have (Python, R etc) on your resume. That is critical.

Second,

If you want to trade options, read the literature. I would recommend two books.

This is basically the options trading bible, this is how I got my job.

https://amzn.to/2ZaG5Rw

The Options workbook, this will refine and quiz you on your knowledge. By taking the tests, I showed my boss I was worth keeping.

https://amzn.to/2KCP1v9

Lastly,

I would talk to traders, hit them up on linkedin or I could mentor you, or whatever. It’s super helpful to hear how they speak. Once you speak the language, it’s much easier to get hired.

We do take directional bets but not how you would think.

We don’t bet on the direction of a stock or future price. That’s not our business.

We do take a directional bet on Implied Volatility though. Similar to a stock that is cheap historically, if IV is cheap historically we would probably be long vega, which means we lose money if options get cheaper and make money if options get more expensive.

We also take directional bets on the yield spreads on the U.S. Treasury curve. I.e. 2 year vs 10 year futures. This usually works in tandem with our options position, but nonetheless we are directionally exposed to that yield spread.

Again, my market is U.S. Fixed income, so I am not choosing the market, I work for an organization that trades one complex of products. So 2s 5s 10s 30s. But again, deciding when to be exposed directionally is based on historical data and future expectations, which takes on a host of different viewpoints and timeframes.

Have you read "More Money Than God."?

It's my favorite read.

https://amzn.to/31EOz52

It talks about the main trading strategies that every top fund manager used. And how hedge funds came to be.. it will give you an idea about how you can be more creative than just outright directional bets. Or how you can do it systemically..

Our business model is to show bids and offers in all options on all products in U.S. Treasury Options, we provide liquidity. It’s similar to a casino that provides you a table to play blackjack. I’m not jumping into random markets and placing directional bets.

We make money by collecting the bid/ask spread (edge) throughout the year while we trade with liquidity takers in that specific market.

Our biggest concern is the total position of all of our options in terms of the greeks we are exposed to, delta, gamma, vega, theta..

So, to answer your question, we look at relative value of options. This happens however you want. Like, maybe we look at the Implied Vol spread between 10d and 30d puts on the same expiry. Or maybe we look at the price of Calls on a 30 day options vs a 90 day option based on historical measures and IV spreads. Or maybe like cross product relative value, 5year note straddles vs 10 year note straddles.

So different from what you are doing, we never really expose ourselves too much to a direction, since we hedge our trades immediately. If I buy 100 30delta calls, I sell 30 futures against it. Like, we wouldn’t buy 1 month puts and hope our market goes down. Timing the market is super tricky. If we put on an options spread, we don’t usually hold it until expiration either. Once there’s like a week or less to an option, the greeks become super crazy, the option loses its “optionality” and becomes more of a binary thing.

For example, if I sold a 90day straddle and bought a 30day straddle, same underlying. And I sold it at say 50 ticks. If it moved to 48 ticks after a couple days, I might take profit. Or if based on historicals, I determine this is a great trade, I might keep it on for 20 days until I make more on it. I would cut my losses in a similar fashion.

In terms of signals, pick whatever you want. Technicals definitely provide some guide, or looking at the underlying trade volume against the various prices that have traded, if there are pockets of low volume, possibly that’s a breakout area.

A bet I would recommend is this. I do this on my own account. If I notice that something particularly extreme happened. Like Hong Kong protests sending the S&P 500 down 3 percent on a given day. Right as the market opens I buy a super tight put spread expiring on that day as the event unfolds. Something small. Maybe I bet 100 bucks. But the risk is like bet 100 to make 1500. And since it expires that day. The prices are generally asymmetric to the event unfolding. Math wise, that put spread has a 7.5% chance going in the money, but since I know this is truly a unique event. The odds are really like 25%, so if you do this every time there’s something wacky happening, overtime, it seems to be a winning strategy. You just have to really understand the global backdrop. Super hard.

This strategy began to make sense to me after seeing it done for larger size by hedge fund players in the market.. and after reading this book "Anti-Fragile."

https://amzn.to/2KCcRXD

It explores things in the world that gain in value when there is disorder or extreme randomness. Which is the general thesis for my trade recommendation..

Your trade, to me at least, is more a synthetic arb(?)

A traditional arbitrage opportunity is buying something for 10 and selling 11 immediately riskless. Your exit value is theoretical and based on factors that have variance to them. I'm not saying it doesn't work/make money. It's just not an arb arb in my opinion. That's actually an interesting strategy, a lot of the bigger firms do just that trade, to provide liquidity on futures for large orders/block trades, or for hedging large option trades with minimal slippage.

But anyway, no, my directional IV bet is not an Arb, by definition it's kinda the opposite. When we do cross product IV spreads, that's a synthetic arb, like your strategy. If Notes over Bonds IV ratio (NOB) is .6, we would buy Bond IV and sell Note IV in hopes the ratio compresses to the historical mean of .55 or whatever it is.

Our directional bet on IV is one of the biggest risks we take and it's our job as prop traders/MMs to know how to best position ourselves smarter than the market as a whole (whether to be long or short IV).

We have custom charts on bloomberg based on the ratio of long term historical relative performance, like realized vol on 10s vs 30s. Then we use that long term ratio to normalize a chart on the future spread. It's similar to looking at a yield spread chart, but its in terms of the futures price which is just easier to intake, and it's less noisy then the yield spread chart.

If you just assume mean reversion then it's easy to decide what to do on the spread. Outsize moves like the past 2 weeks are hard to predict and never mean reverted..

We don't have direct spread bets on either, we derive spreads based on the change in gross option deltas per product, due to all the dynamics of warehousing long and short positions on all the strikes in the products.

But yes, we made money.

Bond options on the call side were at a .61 implied NOB..(notes over bonds).. this ranges basically from .65-.45 ..... .65 means that implied vol on the notes is 65 percent of the IV on bonds. Bond calls were super cheap before the move. Since we were long those calls, and the bonds rallied much faster than the rest of the curve, we derived long bond futures faster than we derived short futures from our short calls in the 10s. So from the big movement we derived a spread being long bonds (30s) and short notes (10s) and that was what you needed to make money the past 2 weeks. I think so far it realized a .46 NOB, that's a huge discrepancy in pricing.

Slightly,

The role of a trader day to day technically doesn't require any programming still. You simply monitor all the data and make trading decisions.

But trading is slowly but surely going fully automatic, stocks and futures volume is mostly algo driven. So the same could be said about the derivatives.. its just harder and taking longer to get there.

But when it does get there, what is a trader going to do? Certainly not twiddle his thumbs, so as your free time slowly grows on the job, the expectation is that you can program tools, generate analysis, fine tune the algos etc.

This is taken into account when good firms are hiring. They'd much rather hire someone who they can train to trade but has proven their technical skills at school, vs the other way around.

So in short, you won't get very far or have a stable long term future if you don't adapt and grow a skill set in programming.

Now let me be clear, you don't need to know the hard languages like C sharp and crap, but you need more of the higher level stuff like R and Python.. Python is king in the trading world.

I'd recommend these couple resources if you really want a job in trading..

The most relevant book for Python as it directly relates to trading. Head quants read this.

https://amzn.to/2KQgFDN

This book is part of the U Chicago masters program, they only cover chapter 3 and 4, but if you can master the whole book, you would be invaluable to any firm and could skip a masters. It covers all aspects of time series analysis and forecasting, which is a bottleneck in trading systems.. the talent is just sparse out there. I got through the first 5 chapters..

https://amzn.to/31Q2RQx

Perhaps just google introductory courses in Python and R before diving in.. not sure what to recommend..

I mean, only rarely will I do something that's clearly a "conviction" trade. And I only do it when its so obvious that like when i see the opportunity i literally lose my breath, because the price i see is so incorrect i just instinctively act on it.

For example, mid week last week 2s10s yield spread went negative for the 1st time since 2007, that's the main recession barometer.. and we started to hit new all time contract highs.. in those moments the market tends to move parabolic because algo driven liquidity all dries up in parallel.. and I was on the European shift before smart money with real exposure wakes up.. so I just started buying calls outright, paying the offer, vol was super expensive to start with and we were at historical future highs,, so really the bet i made was stupid, but i just new we were about to gap higher, and 2 hours later we rallied another 2 standard deviations and vol went up 20% to insane levels.

So yea those are the moments when I just pull the trigger and bet, cuz again like the feeling is so palpable. And its different than betting on a direction on futures, because you can see how the market prices option skew and where theres clear pain points in the options, so not only did i buy calls, i bought calls that were long vega prime (rate of value increase, increases as you raise vol), and i bought the calls where i knew the market was short.. so there was pent up demand for people to buy with me to cover their shorts.. and the U.S. is like the only place in the world with positive real interets rates lol so people need our yield.

But generally for long or short IV positioning, its based on historicals in the IV and realized vol, like, is realized trending up or down (maybe looking at rolling 10 or 20d range vol or close close vol), where is current IV, how big is the discrepancy there, what data is coming out over the term of the contracts, and is the market pricing the data correctly ( we have a sheet that tells us the weight that is on each piece of data, like unemployment number is worth 1.8 days historically or something, and then we back into what that should be in black scholes vol terms)

So we do have these gauges we look at, but also, it's just having an understanding of the global backdrop, I work with guys who've been trading 20 + years, so we just debate all day whats going to happen and position accordingly. And yea we do look at technicals as an informal guide to supports and resistance when hedging with futures.

But for quantitative stuff we don't have anything too sophisticated right now to be honest, these bigger firms have pretty robust vol forecasting methods, like looking at 5 min range bars and forecasting 1 week or 1 month vol. So near term options tend to be priced more efficiently than they were 10 years ago as a result, but like ok, ill just line up my markets with them.. But yea we don't have any crazy quant automated systems or anything. Those are super hard problems. Like its not a trivial thing developing an automated gamma hedging strategy using quant analysis, because simply put the past isnt the future. Like if we relied on some systemic strategy over the past 2 weeks, we would have gotten smoked. A lot of the quant shit happens on the low latency front and is more about gaming that system, squeezing down the time window to generate alpha/high sharpe ratios.. theres not some like magic 8 ball out there (unless you're renaissance technologies..)

Haha this is fun for me. And low key, like, solving problems is what the universe needs.

As far back as we think is relevant. When moves are large, snap back to other large moves, when the ranges are smaller, look less far back.. I guess. There's no hard and fast rule, other firms may have more concrete rules with this. We don't model it explicitly, the incentive isn't there yet for us. Firms with more robust trading systems must use like GARCH modeling or some other academic way to model the vol, where there's like a persistent term with X lookback window and a high frequency term with Y lookback window and Z other factors. Because yes, there is a herd mentality in terms of pricing the options term structure, like how vega options are priced relative to gamma options, but it has to originate from somewhere. My guess is firms like Citadel have cross desk resources, like their Delta 1 futures algo team commoditizes their vol forecasting and lets the options desks use their material.. is my best guess. And again the incentive isn't there for us because we just fit to market, collect edge, and determine where there are mispricings based on events of the contract term. Our indirect model would be more of our spreadsheets we have built up that allows to do a human version of the GARCH, just by memorizing the trends on our sheets. Idk if that makes sense..

And we capture the realized vol by gamma hedging.. this is the main thought behind options pricing and how the black scholes model was developed.. if the option was perfectly priced, as the underlying moved around until expiration, and you continuously hedge your deltas (long calls, futures rally, sell your excess option deltas that you derive.. for example) then at expiration, the money you made/lost on your continuous hedging should equal the money you lost/gained on your option. Kinda fascinating.

But yes, we capture the moves by hedging, not continuously though obv, your hedging frequency shouldn't matter in theory, but we like to lean long gamma and wait till larger than average moves to hedge, gamma hedging can be fun because you are forced to day trade the futures..

Special gift haha, I like you.

But in all honesty I think I'm just a little crazy/delusional and apparently smart to boot.

Lol, but nah there's definitely some identifiers from my life that kept me convinced. In college I didn't go to a single class or do anything school related (I definitely had some issues hah), I would only study a couple days before finals and Ace most of them.. crazy. Like, I taught myself Econometrics, the whole semester, in 16 hours. I also have pretty good memory.. Growing up I could recite every statistic about every car on the road (was obsessed). And then, as I mentioned in the podcast, playing Halo as a teen, I literally never lost. So those 3 things, quick learning, deep memory, success in high pressure situations, is how I knew. I'm really just explaining this to you, for me, it was a Fact that I was going to be a trader, you couldn't tell me otherwise. So here I am :)

My son can do whatever he wants aslong as he's humble, kind, and intellectually curious. Otherwise imma beat him.

Depending on how short, our margin call would probably wipe us out the next day. Startups have less capital. But we would [almost] never be directly exposed.

If I was at my previous firm, we actually experienced that scenario on an extremely large scale. Janet Yellen paused on raising rates in most of 2016 which no one expected. We were short 100s of thousands of front month Eurodollar calls (the main bets for fed positioning). So we basically experienced the 30 year move you're talking about but it was on 3 month libor instead, and the move was even bigger... most senior level people got fired and the losses were.. unfathomable. The firm was so big though where it didn't matter at the end of the day. That was cool in own right.

I've never heard of that tbh

check out quantopian.com for strategy development.

I don't know much about the bank side of things. "Derivates" could be any derivative, like OTC, Swaps etc. and I'm pretty sure Dodd-Frank got rid of bank prop trading, so its unclear what their business model would be.