I found out in the risk navigator (IB) you can do Settings -> Reference Margin Type -> PMRGN which appears to put it into portfolio margin mode. I have been considering switching to this but wanted to do some scenario analysis first to get feel for it. With a delta neutral stock + put combo, $100k appears to control over $10M in stocks (100:1 leverage). Obviously that's inadvisable, but it seems that it has way lower margin requirements than I had expected. Is anyone that is using portfolio margin finding extremely low margin requirement like this on a delta neutral position?
Portfolio margin is a risk based system. So if you are long a stock and hedged with a combo, your margin will be very low.
Long stock with a put is a synthetic call. So yeah, one could go out and put all their money into 0.10 delta calls and achieve a ton of leverage. Way more than being long the underlying. The shorter the duration, the more the leverage.
As you’ve created a synthetic call the margin requirement would be just the price that you would have had to pay for the equivalent call (IB matches your stock and option positions even if you open them at different times).
It appears that individual stocks run the risk array from -30% to +30%. I found that a short call works a little better than a long put. It is so far ITM that its time value is negligible and its delta is unaffected at 30% so that it is calculated to have zero directional risk. The high capital requirement for the option doesn't matter here only the margin - $37500 margin on 1000 contracts is actually the lowest it can ever go based on a FINRA rule for portfolio margin. If I use the 8 call instead of the 5, it only adds about $300 to that. Further and it starts to increase rapidly. The actual max leverage ratio varies with the price of the stock relative to this minimum margin. Now, my understanding is that for this example I'm going to be paying the IB margin rate on about 493k @ 3.4%, while I earn about 12% on the stock which pays monthly dividends of about 1%. This is of course amplified enormously relative to the margin required. Even if I had long options and was paying 3.4% on the entire position, it would only eat about half the dividend. It looks like the return on margin is higher than 200% yearly. Also dividend risk (if lowered 10-20% which could happen for AGNC) mainly affects the price of the stock which is fully hedged, so it would just lower the return a bit. Did I make some kind of mistake here??
Your short calls will be assigned prior to the first dividend, closing the entire position. There's no free money here. Another thing to be aware of is that IB has a 30x gross leverage cap, which can be more limiting than margin if you do stuff like this.
First, a 1% dividend results in a drop in stock price of 1% on the ex date. The company has to actually be growing at a rate the justifies a 12% growth in stock price to counteract that 12% drop over the course of the year from a 12% dividend rate. This perception that a stock dividend works like a bond coupon is common but entirely incorrect and is the major flaw in your scenario. Second, a 12% dividend is huge. That's more than 6 times the Treasury rate. If you buy bonds with that yield they're deep in junk bond territory. That rate is representative of the extreme risk in that stock, which isn't necessary bad, I'd just caution to make sure you get the concept of risk adjusted return before making a call based on the big shiny number.
Yeah it occurred to me that long puts would have to be used for the delta instead; This is a REIT which is why the high div yield, as are the others I was examining. REM is a broader index of REITS yielding 9% in the case I am worried about company specific risk and might be a better way to go, but I do prefer to use LEAPS so I don't have to roll very often. AGNC and NRZ have them, REM only goes out to ~200 days. I did discover that the risk algo won't examine correlation of individual securities (though as I understand it it will use index beta risk). For example I have a long REML with long puts on REM because delta on one is very close to delta on another in practice. When it comes to stock+option combos, it must be the same underlying, but that is not surprising. There are specific cases where long+short on indexes can "cancel out" risk, but not for what I'm doing. One thing I could not find is how to split out the "regular" time value from puts when dividends are involved. I ended up using the CBOE calculator and changing dividends to zero to compare. IB doesn't seem to be able to split it (but ToS couldn't either, at least I didn't see how to). IB can also only display time value at the bid so I am constantly adding the half spread to the displayed value manually; normally I want to know the time value at the mid.
Why not just buy the 17 call? The value of the dividends you'll receive is embedded in the price of the put, and you mostly lose the beneficial tax treatment of the stock dividends if you hedge with a put. Also, the cost of carry priced into the call will be better than the 3.4% you'll pay IB to hold a synthetic call (though you could do better by selling SPX box spreads).