Thanks to all for replying. I see the point on a slow move down, that makes good sense. Potentially, I see doing something like this as part of a diversified long strategy...this would be on the "insurance" end of the spectrum.
Absolutely. But if you are willing to spend 6% per year for insurance, buying options with longer expiration will be much more effective. For example, Sep18 200 puts cost around 2.70. You can buy 20 of those puts per 100k portfolio, and they will provide much better protection. If SPY falls to 200 (20%) and IV increases by just 10 points (this is conservative estimate), those puts will gain around $40k. And you can roll them 2-3 months before they expire. More effective, less slippage, less commissions. But again, I believe paying 6% for insurance every year is way too much.
But what if the SPY went up 18% in 9 months...just for the sake of argument. Had you done that last year, your strike would have been 170 or 175. You could buy 1/4 of the contacts quarterly, and that would help, but it's still not particularly effective.
The no/low cost collar is viable if you're willing to give up the upside (the short call). This converts the long position into an equivalent vertical. If the underlying collapses and you believe it can recover, you can roll the collar down (or just BTC the puts), lowering your cost basis. There will be losses but they'll be manageable.
Yup. I've been playing with VIX ratio spreads trying to get around that. The problem with them is you can only pick up the early move, and there's no bearish equivalent (though, that's somewhat mitigated by the lack of "panic buying"). Also, you can only purchase on low VIX of you're trying to reap the benefits of long term hedging with it. Very inexpensive for what it does, but not very valuable for capturing a big move since it loses sensitivity the higher the VIX goes. I'm fairly certain on it's own it's a losing strategy, but not by much. But in the context of a cash injection when buy opportunities arise, it might be profitable overall. It's really not something I'd recommend for most people, and definitely not useful for the question in the thread, but kinda cool nevertheless.