At the end of the day the market is a sandbox and you can develop your view of it as you please. Volatility is seen as a common proxy for risk. But it’s not static.. and arguably we as a collective market can’t avoid understating risk of extreme outliers in order for everything to run somewhat properly the other 99% of the time. So the distribution curve has fat tails.. this is commonly known. Kevin gave you a pretty solid answer of financial orthodoxy. You pretty much restated your question.. just apply what he said. To make greater returns you can either operate a better strategy or sit on more risk. With SPY vols we already suffered a very statistically unlikely DD of 50%+ in recent years. If you want to hold that vol 1.5x’d you’re going to get paid more. But how confident are you that if vol is 1.5x, the worst tail event is going to perfectly follow that ratio? If your thesis is wrong you’re going to start edging up to some real insolvency risk. Which is fine but at least be clear on what risks you’re holding. There’s a few much more low-hanging fruits to harvest before you go the route of brute force leverage. Lower yielding assists have been found to have consistently higher risk-adj returns over equities.. perhaps due in part to leverage aversion as we just discussed and also since more of your liquidity will still be there when you might need it most.. i.e. a lot less volatility drag to realize. Also holding less correlated assets simultaneously and rebalancing them smooths global portfolio vol without dampening returns to the same magnitude. A lot of these foundational concepts have been very well discussed over recent decades under Modern Portfolio Theory. If you understand the implications of the above, some ways to squeeze more yield out of your investments will be obvious. It’s the next wall after optimization via a backbone of broad statistics that’s got me stuck at the moment. I feel like there’s got to be the next step of iterative refinement from this point just like there is from 100% long equities to MPT.. but I still don’t know quite where to look.
Appreciate your feedback sir, just like I appreciate Kevin's. I am not challenging CAPM, in fact I use it. Same with fat tail, it is a two edge sword. I am beating my brain out trying to understand risk/volatility, randomness and the behavior of the market.... Regards,
I was probably wrong and you were right in your intuition on long-term risk in SPX vs RUT. I assumed that, within an asset class, other measures of risk would be proportional to vol. This appears to not be the case: the market is pricing in much less (proportionally) downside tail risk in RUT than in SPX. Probably not for a long term investor like you. Market risk premia generally compensate sellers of insurance against downside tail events. So a risk metric that measures the chance (risk) of those extreme negative events is probably what you want. Risk premium in a given instrument tends to correlate highly with measures of downside tailweight -- google "tailweight measure." Or you could assess the risk premia more directly by looking at the ratio of, for example, far OTM put vol to ATM vol at the timeframe you are interested in for the instrument you are planning to invest in. I was wrong. I was assuming that downside tailweight in SPX and RUT were proportional to their respective vols. Not only is this incorrect, but it is more incorrect the farther out in time you go: Sym Expiry atmIV call25dIV put25dIV RUT ,20181116, .196593, .174547, .226807 SPX ,20181116, .135073, .113461, .166619 RUT ,20191220, .178938, .142705, .224632 SPX ,20191220, .154839, .111784, .20676 * as 20181109 close As a case in point, the financial crisis drawdown for SPX was ~56% (ln -0.82) and for RUT ~60% (ln -0.92), or about 12% larger for RUT instead of the 50% larger number that the risk proportionate to vol assumption implies. In short, if you have a 30 year investment horizon and you expect that RUT will return more than SPX return * 1.12, and you believe that the market price for tail risk reflects your true risk, then you should choose RUT over SPY. On the other hand, neither is likely optimal. A passive diversified basket of lowly correlated assets might be a more reasonable choice.
Thank you again as your comment is what I have been looking for: I think for my short term option trades, volatility is as good as it gets for defining risk and I better pay attention to it. Longer term, risks of an asset/asset class may be affected by additional components other than volatility: Something non random is affecting long term stock prices so in that case volatility is not exactly risk?
There are some common measures of risk including Standard Deviation, Beta, value at risk(VaR), and Conditional Value at risk(CVaR). Risk management simply accesses the level of risk and then steps are taken to minimize the level of risk,taking into account the potential return and your risk appetite. A 'Guaranteed Stop' is one way to make sure you never lose more than you can afford even though the market ‘gaps’ during a volatility period, one won’t lose more than the amount of the stop.
%% I can solve part of your problem; markets are not random+ you/anybody will never completely understand it. Same way with electric power+ SPY. SPY has done well over long term- I would not sit around in the dark until I completely understand both.I don't auto assume a position is too big if my guts bother me- I do more research on it. But if my brains felt like''beating my brains out'' , frequently, make changes...…………………………………………………………………. Actually I hate the way small caps outperforms SPY sometimes , but not this year.:LOL, [So I made some changes...….]
Return on risk in many options spreads. Not all, but that mostly doesn't apply to calendars and European verticles.