Back to Community
IB Effectively Made VXX/UVXY un-tradeable with new margin requirements

Anyone else in the US suffering because of their new margin requirements? Basically you can only trade with 50% of your cash if you're short VXX/UVXY. Very unfortunate as short VXX/UVXY with timing is a no brainer.

48 responses

i heard the new margin terms apply from this weekend.

They're live right now. I tried to short my total cash balance and it said I did not have enough to cover my margin requirements.

Hopefully if/when volatility comes back to the market they alleviate these margins b/c they are just plain silly. They make you hold a margin as if the VIX = 18.

I understand they're trying to help people not lose all their money, but they really put a damper on my party. I was up over 60% YTD, but now I can't trade my strategy and I'm definitely not buying XIV - nowhere near the same as shorting vxx.

at this point I believe XIV is the second best choice considering the margin terms has no effect

Yes, my algos have been impacted by this change too. I called up IB and talked to them, asking if this was a temporary measure. They stated that it is not temporary, they have no plans on changing these new margin requirements.

Also, the shorting is not 50% for securities like TVIX, it is less than that. I think it is around 35% - it is really ridiculous!

It would be great if Quantopian would support some other brokers, like etrade, tdameritrade, schwab etc.

But I dont think Quantopian will be supporting us in this venture, I dont think they like VIX traders at all. Almost every VIX algo needs spot, 1 month and 2 month vix data, but they still do not support this.
So we end up having to resort to using quandl or some other third party service, which by definition has to be stale data (i.e. data from yesterday because it is retrieved using fetch_csv).

Having said that, when you shut down your platform for a whole day, I guess you really have MUCH bigger problems to deal with first.

This is frustrating - but I can manually trade my strategy - I would just rather have quantopian do it for me.

Mohammad, do you know margin requirements on etrade/tdameritrade/schwab?

I have a personal backtest of XIV/VXX VIX/V1/V2 from 2004. My short VXX strategy (although it doesn't account for borrowing/transaction costs) starting with $30,000 in April 2004 ends with $59M and yes that is even after the 15% drawdown from last week - it was at $72M! But when I switch out to go long XIV instead of short VXX it ends with $8M - that is a huge difference!

I do not know what the margin requirements are for the other platforms, but I have heard that their margin interest rates are much worse than IB (i.e. if you are using leverage).

Although it is tempting to compare the end result, it is actually a poor comparison metric. I prefer to use CAGR - something Quantopian doesn't show.

The reason is that the interest you pay for shorting is not factored into the Quantopian results.

So let's say shorting VXX costs you 10% interest per year - you can just take the CAGR and minus 10 to get a CAGR that should be a bit more accurate.
Then you can compare it to the CAGR for trading with XIV (since you would be going long XIV and there would be no shorting interest).

Mohammad, your input in CAGR is very true - I should focus more on that than the end sum.

Long XIV came in @ 50.88%
Short VXX came in @ 74.95% - 10% borrowing = 64.95%

So net a difference of ~14% shorting vs going long XIV

That doesn't sound surprising - I have seen similar differences with my own backtesting.

My biggest issue is that the drawdown is also different, in my algos using XIV has a higher drawdown and lower CAGR. The advantage is that you have defined risk, whereas shorting is theoretically undefined risk.

IB is smart.

The initial margins for short UVXY have increased to 300% and have stayed there for a few days now.

Yes, and IB has stated this is the new norm. So you can only short about 33% of your account, no more.

When you say "untradeable" do you mean your algo has become unprofitable with this new policy? To make the best of the situation, couldn't you simultaneously short VXX 0.5 and long XIV 0.5 (or whatever ratio keeps you within margin requirements)?

Depending on how the margin works - you would be buying the XIV on "margin" or borrowed cash, no?

Because the margin requirement zaps your buying power.

This is very unfortunate, but so many traders have exploited VXX & UVXY especially recently. They are so easy to trade profitably, it's almost to good to be true and I am honestly wondering how long the party will last. With IB making this move, I wonder if this may spur other brokerages.

Im thinking of switching brokerages from IB cause this UVXY margin increase is just atrocious. Which brokerages have more lenient UVXY requirements?

I basically moved away from IB. They blocked my account from trading VIX ETFs for a while claiming that the risk was too high.

When my broker decides to police the markets, I know it's time to move (even though I had been with them for many, many years).

I have started using Ally. They have much better margins for shorting VIX related ETFs. Depending on your account size and strategy, their fees may be lower too (I suspect for smaller accounts their fees will be higher).

And another HUGE bonus... they have a really simple REST API if you want to build your own algo on top of it. It is so simple it puts every other broker API to shame. (But they don't have paper accounts to test their API, which is... REALLY bad)

Thanks Mohamed! Do i need to be an ALLY customer to have access to their brokerage account?

Also do they usually have locates for UVXY and VIX, i found that etrade charged way too much money. Anyone have any luck on Tradestation?

I signed up with an Ally Invest account. I have only just started using them, but I did not run into any issues shorting UVXY.

Mohammed, What specifically did IB forbid you to trade? I understand the higher margin requirements on shorting VXX. Frankly, they need to cover the risk of VXX tripling or quadrupling in value. In a 1929 like crash that would actually happen, so a 300% margin requirement is not so crazy.

Are you saying they have directed you not to go long XIV?

My understanding of buying XIV is that it is akin to buying a long term put on VXX. Because VXX suffers from contango degradation greater than the time loss you always normally suffer from holding a put, with VIX you are not suffering from net time erosion if volatility remains steady. I know there is some kind of clause in the XIV whereby the value could go to zero under extreme circumstances, but I confess I have not YET dug deeply into what those circumstances would be. I will do that before taking any large positions in XIV.

In any case, IB would not be risking more than 100% of value of XIV, and consequently should not require higher than normal margin. Are they doing so?


IB forbid me from increasing margin on ANY volatility related securities. So I was not able to short TVIX or UVXY. I was also not able to buy XIV. I was not able to do anything at all related to shorting volatility.

They basically locked my account, because they claimed that the risk of a volatility spike was high.

Note that I was shorting TVIX and only used a fraction of my account to do so. But they still locked my account from shorting volatility in any way.

I have an new policy now... when your broker starts to forecast the future and police the markets, it's time to switch brokers.

Thanks for responding. Good luck with the new broker. I fear that IB is the canary in the gold mine, and all the others will follow suit shortly.

I have been trading XIV actively , as much as $100k (for my bigger $million accounts) and never for more than 10% of portfolio. But I'm even uneasy with that since XIV could go to zero on a stock market drop of 5% or more in a day. Rare events yes, but I would be shocked if we don't have at least one over the next 8 years. See the link here for explanation: vix shock. Frankly, I have not studied this in depth, and the analysis may be flawed, but I would rather err on the side of caution. I will continue to trade the XIV but with the assumption that in a big market move it will go to zero and be "closed" at that point by the ETF. So there will be 100% losses on XIV positions.

Serge, the point is not that volatility trading is not risky - everyone knows it is.

Weather it is risky or not doesn't matter, that FACT is that by not buying XIV 12 months ago, you missed out on OVER 100% gain. This is a fact.

I'm not saying you should buy XIV. But IB told me that I cannot buy it. That is because THEY deemed the risk as too high.

The fact that trading occurs at all is because different people have different assessments of risk and value. When a someone tries to force THEIR opinion of risk onto you - that is a very bad thing.

BTW most people that trade volatility products have their trades hedged, and I do too.

Wherever you sign up with a broker they are required to give you a survey asking you what your risk tolerance is. Perhaps if you update your survey they will be less restrictive with you. Could possibly be the issue. Otherwise that's super strange.

If you have no need for shorting, I recommend Robinhood as it is free and there appears to be no limitation on trading XIV (unless you want to trade it on margin, but any medium to high volatility security will have restrictive margin maintenance requirements anywhere you go). (PM me if you want a referral link, and they'll give us some free stock.)

Viridian, I don't think that is the issue. Since I have a Registered Investment Advisor account with IB, I have a feeling they know that I am not a newbie.

Also, I don't consider Robinhood a proper broker, they would not be suitable for my trading.

hi guys, my another broker in Hong Kong also forbid shorting UVXY/TXIV. any alternative solutions?

Easiest way is to use options. Do a synthetic short i.e. sell a call and buy a put (at the money).

Hey Mohammad,
how far out do you buy your synthetic shorts for? since options expire

It doesn't really matter, but generally if you try to stretch it more than a couple of months, the spread means that you will have to pay a debit for the options.

If you buy the next monthly, you can usually do a synthetic short at no cost (because of put-call parity). You can always roll the options once they get near expiry.

are you doing these synthetic trades yourself?

you sound like you know what you're talking about. Are you some sort of trainer? I've traded for some time now and looking to get better. I'd love to connect with you if that's okay.

I do short volatility using options with various strategies, but I don't do straight synthetic short (because the risk is too high).

I trade for myself - I don't do any training, but if there is anything I can help you with feel free to connect with me on Linkedin.

I'm starting to run an intra-day VXX/UVXY/SVXY strategy and am looking at a similar problem with margin requirements, especially after Feb 28. VXX is 1:1 now at least on IB.

Since I'm only holding intra-day I'm considering a synthetic short position on the VXX and UVXY - perhaps with weekly options. The system closes out before the bell, so the main concern would be the two spreads. Mohammad - you seem to have a lot of pertinent experience - can you offer any thoughts?

VXX options are much more liquid, meaning the spreads are tighter. Everything you can do with UVXY can be done with VXX, so you might want to stick to VXX and adjust your calculations appropriately.

Also, to do a synthetic short (i.e. sell a naked call), you will need an account with $100K+. If you dont have that, you could just buy a far out-of-the-money call for a few pennies so that it is technically not naked.

If you have a large account, you might want to use a different broker. IB option contracts are $0.75 each! If you are trading actively this can end up being a lot in commissions. I was paying over $4,000 a month in commissions last year, until I decided I needed something cheaper.
There are many brokers that will do contracts for much cheaper than $0.75, the average is about $0.50 but my fees are way lower than that. If you want more details PM me.

Thanks Mohammad. Account size isn't the problem, but you're right about the commissions. How do you feel about TS? I would have to write or hire someone to write a port for NinjaTrader or python but it would probably be paying for itself in the first month. I'll contact you privately to discuss this further.

@Mohammad - I wanted to circle back to your response as you talked about selling a naked call whereas I was talking about a synthetic short = selling an ATM call and buying an ATM put, both at the same strike. Aside from retail commission, per your own experience do you see any major downside in using such an option construct in place of a VXX or UVXY short position? Thanks in advance.

If you are holding for more than a few days, synthetic short will be cheaper than a real short because you wont have to pay the borrowing costs (borrowing VXX or UVXY is really expensive!).

The only real world challenge is this... when a vol spike occurs, as the call goes more ITM and delta approaches 1, you will be assigned on the call (because there is not much extrinsic value left). As long as you have margin to close out your shares and place another synthetic short, this is not a problem. But if you trade aggressively, you will see you lose a lot of money, at the worst possible time.

Similar scenario plays out with a real short. After a vol spike your broker will call you and tell you to close your short VXX trade. Why? Because VXX has gone from easy-to-borrow to hard-to-borrow - again, at the worst possible time for you :)

If you want to short vol, you should consider an option strategy that limits your loss.

You could always open a long put spread. You could buy an ATM put and sell a PUT OTM - this will be a debit, but doesn't leave you open to catastrophic losses.

Yes. My preference is actually a backspread... sell 1 put ATM, buy 2 puts ITM. You can make this theta positive and your profits are not capped but your losses are.

Awesome point Mohammad - I forgot about backspread puts. How far ITM do you go, or is your preference?

Also - how far out do you normally open these?

That's some excellent input. Like Tyler I was also considering a simple bear put spread, it's actually the safest route I recommend to my own clients. Mohammad, you're right about the surge in delta probably triggering an assignment, not really what I would want to deal with.

I don't trade aggressively, position size should be around 1% of capital, although margin requirements makes that more tricky with those ETPs. Also important to note that I only hold intra-day, the system always closes out before the EOS.

I'm still testing various things, but I try to make it such that I don't pay premium for my puts. Which means you are often 20% to 30% ITM for 2 month options. 2 months allows you to weather most storms.
If you opt for shorter duration say 2 weeks, you can often go less than 10% ITM and still not pay premium.
And if you don't mind paying premium... the world is your oyster :)

Remember that the payout plays against you with a backspread - you lose twice as fast as you gain. So winning 3 trades and losing 3 trades means you are very far away from breakeven - which makes duration (i.e. weathering of vol spikes) very important.

Mohammad, as I mentioned above - I'm only in the market for a few hours at most. What do you think would be the most optimal configuration via an option strategy?

If I were personally shorting vol for a few hours and not holding overnight, I would buy a put 7% or more ITM, shortest duration possible. This will keep your theta decay and gamma losses low while keeping the number of contracts you can buy reasonably high (allowing for higher leverage).

Remember you have to get over 66% wins to make any profit from backspreads - with very short time spans that is hard. Buying a single put would make more sense in this scenario; you only need to be right 50% of the time to start making money.

Mohammad, regarding the required win ratio for a backspread - I assume that is due to the fact that you lose twice as fast as you gain?

Yes, although I am making a lot of assumptions, the general gist is that because you lose twice as fast, you need to make twice as many wins as losses.

@Mohammad, simple ITM puts with sufficient delta actually were my initial instincts, perhaps weekly options if available and if the spreads are decent. This could work particularly well during the last two days of expiration as options sort of turn into little mini futures contracts.

Reg. using ITM front month or weekly puts for short term campaigns - a friend of mine who works at a larger firm mentioned this to me yesterday:

If you're not up on day trading options, a refresher of the CBOE's market maker rules for spreads is a good primer. The rules for spreads are much tighter for > ITM options.. i.e. once a spread goes OTM, they can widen the spreads substantially. As long at the option stays ITM when it moves against you and hits the > stop loss, you're good. Assuming you're going long puts.

Very valuable consideration to go deeper ITM to avoid widening of spreads. My initial stop loss thus should define the strike before where I would expect spreads to widen.