Are You Reading Implied Volatility Wrong?

Are you looking at implied volatility the right way? Mat Cashman of OCC joins Jeff Praissman to explain why an IV number doesn’t mean much without context, and how comparing implied volatility with historical volatility can give traders a more useful framework for understanding options pricing.

Summary – IBKR Podcasts Ep. 423

The following is a summary of a live audio recording and may contain errors in spelling or grammar. Although IBKR has edited for clarity no material changes have been made.

Jeff Praissman

Hi everyone, this is Jeff Praissman with Interactive Brokers. It’s my pleasure to welcome back to the IBKR Podcast Studio, Mat Cashman from the OCC. Hey Mat, how are you? 

Mat Cashman

Oh, I’m doing great. How are you today, Jeff? 

Jeff Praissman

I’m great. Love having you in the studio. And for our listeners, even before we get started, you can always find great material from Mat and OCC on our website, interactivebrokers.com. Click on Education. On their website, occ.com, or the OIC website, maybe more importantly, optionseducation.org. 

So today we’re gonna talk about one of my favorite subjects. We’re gonna talk about implied volatility. It’s one of those terms that option traders use constantly. They’re throwing it out there. If you trade options or you talk to someone about options, you’ll hear them say volatility is high or volatility’s low or the options are pricing a lot of movement. 

But what does an implied volatility number actually tell you? 

Mat Cashman

That’s a great question. What does it actually tell you? The first thing I’d say is that the number itself, just the actual metric, the number that people talk about, usually doesn’t tell you enough. It’s not a perfect way to describe what’s going on. And so if you tell me that a stock’s options have 60% implied volatility and ask me whether or not that seems high or low, my next question is probably gonna be: compared to what? Because that’s what really matters here. The 60% number that you’re looking at might be an elevated implied volatility for one underlying, and it might be completely normal for a different underlying. And that’s really the framework that I think we wanna get into today when we’re talking about it—not just looking at an implied volatility number, but understanding what that number really means in a meaningful context that you build around it. 

Jeff Praissman

Yeah, and I think it probably makes sense. We got a lot of experienced option traders listening to our podcast. We get some novice ones as well. Mat, before we get into these comparisons, probably makes sense just to define what exactly we’re talking about when we say implied volatility. 

Mat Cashman

Yeah. Another great question and a great place to start. Implied volatility is really the volatility assumption that’s embedded in the market price of an option. So that’s a lot of fancy words to basically describe, importantly, implied volatility is an options metric, just like any other options metric, but it’s associated with price. And so we observe the price of an option in the marketplace very intuitively, right? When people look at option prices, they understand what that means, but there’s an implied volatility assumption behind that that is consistent and commensurate with that price, given all the other inputs that are being built into that model. Also very importantly, implied volatility is a forward-looking metric. So it reflects the amount of future underlying volatility that’s currently being priced into those options. Historical volatility is different. It’s backward-looking. It measures how much the actual underlying is moving around or has moved around over some previous period of time. So right away, we have two very different pieces of information that are kind of pointing in different directions. Historical volatility—what the underlying has actually been doing over N number of days, however long you wanna make that—versus implied volatility, which is a forward-looking metric thinking about how much movement we might see in the future. 

And that’s the useful comparison between those two things. 

Jeff Praissman

So just to simplify, historical volatility is what happened in the past, and implied volatility is what the options market is pricing going forward. 

Mat Cashman

Yes, that’s basically the crux of all kind of options trading because what you have in front of you is the previous amount of information, and then you also have a forecast going forward, and you have to make a decision about whether or not you agree with that forecast. 

I sometimes think about historical volatility as the scoreboard and implied volatility as the forecast of what’s gonna happen. So historical volatility measures something that’s already happened. Implied volatility reflects the current expectation of what might happen going forward, but neither number exists in a vacuum. And that’s kind of the important part about this conversation. If an option’s implied volatility is above the actual comparable historical volatility for the previous number of days, that doesn’t automatically tell you that the option is mispriced. And if an implied volatility is actually below that historical volatility number, that also does not automatically mean something is wrong there either. There may be a perfectly logical reason for that relationship and for those levels to be different—the implied volatility versus the historical volatility. And that’s why I like thinking about this through a relative value framework instead of immediately trying to attach some label as to whether or not something is high or low just based on the absolute implied volatility number alone. 

That number doesn’t really tell you a whole lot. And that’s why I wanna have this conversation. 

Jeff Praissman

All right, so let’s just say someone pulls up an option chain and they see the options trading at 80% implied volatility. Their immediate reaction might be, “Oh, that seems high.” But from what you’re saying, that’s not really how they should look at it. 

Mat Cashman

Yeah. The answer to that is maybe it is high, but maybe it’s not, right? The real answer is we don’t know that yet. Suppose the options on one underlying normally trade somewhere between 15% and 25% implied volatility. That’s where they like to live. If those options are suddenly 80% vol, something has obviously changed. But suppose the options on another underlying routinely trade between 70% and 100%, and 80 vol there might not be particularly remarkable or different than what it is every day. It’s the same implied volatility number, but the context is different because you’re building it into two different frameworks: 15% to 25% implied volatility being normal and 70% to 100% being normal. And so that 80% number looks completely different based on the context. It’s a little bit like looking at stock prices. I like to use this as an example. One stock trades at 10 bucks and another stock is trading at 500 bucks. You wouldn’t necessarily look at those two and conclude that the $10 stock is necessarily cheap and the $500 stock is necessarily expensive simply because of their absolute prices. You need some reference points before the numbers become meaningful or can tell you anything. And so implied volatility works very much in that same way. And I want people to start to look at it in that way as well. 

Jeff Praissman

That’s a great analogy with the stock prices, Mat. The $10 one could be overpriced, actually, when you think about it. So how does a trader know what reference point to start with then? 

Mat Cashman

That is a good question also. 

Jeff Praissman

Mat, full of good questions. 

Mat Cashman

You’re full of good questions today, Jeff. I mean, it’s kind of instructive to say, “Okay, well, now that we’re at this point, where do we go from here?” And that’s kind of what we wanna think about when we’re thinking about a topic that’s as complex as this is. One really straightforward approach is comparing those two metrics that I talked about earlier, right? Implied volatility, which is the forward-looking metric, with historical volatility, which is the backward-looking metric, but think about them over a similar period of time. And one way to express that relationship of the implied volatility number and the historical volatility number is as a ratio. 

So think about it like a perfectly normal fraction that you would think about any other—one-quarter or two-fifths or whatever, right? You think about it and you say, “Take the implied volatility number that I’m looking at and divide it by the historical volatility number.” And so the implied volatility number is the numerator, the historical volatility number is the denominator. And that can really sound way more complicated than it is, but what you’re essentially asking when you boil it down to this very simple ratio is: How much future volatility is currently reflected in these option prices relative to the amount of volatility that the stock has shown over the last number of days, or how much it’s been delivering? And that’s the basic idea. Once you start looking at volatility as a relationship rather than just this isolated number, right, like 60% or 80% or 100%, you can make comparisons that weren’t necessarily obvious from just looking at their absolute values. And that’s where this relative value context starts to really become important. 

Jeff Praissman

So, Mat, I think an example might be helpful for our listeners. If you can just maybe walk through an example with numbers, explaining the relationship you just talked about. 

Mat Cashman

Yeah, that’s a good idea. Let’s say options on Stock A have an implied volatility of about 20%. So the stock’s 30-day historical volatility, how much it has been moving, is delivering about 15% over the last 30 days. That gives you an implied-to-historical volatility ratio of roughly about 1.35. And that’s the 30-day options divided by the 30-day historical. That’s all that is. 

Now, let’s take another Stock B. Its 30-day options have an implied vol of like 58, and the underlying 30-day historical volatility is roughly about 43%. Those are completely different absolute volatility levels from Stock A to Stock B, but that ratio is also, guess what? Right around 1.35. 

So if you look just at those implied volatility numbers, you might say those two stocks don’t have anything to do with each other from the way that the options are priced. One of them’s a 20 vol and one of them’s a 58 vol. But when you build in the idea that there’s also historical volatility that’s influencing where those vols are, you get some similar amount of ratio between them. So if you looked at just the implied volatility numbers, you’d think they were nothing alike. One set of options is 20, one set of options is almost 60, but the relative amount of movement in each underlying that has actually been delivering shows that it’s reflecting a very similar relationship between the implied vol and the historical vol for both of those stocks, Stock A and Stock B. That’s information you don’t necessarily see when you focus only on the implied volatility number, right? 

Jeff Praissman

Yeah, so Mat, in both examples, the implied volatility was higher than the historical volatility, giving it, like you said, a ratio of 1.35. So am I understanding correctly that if the ratio’s over one, then the options are somehow overvalued, or is that not what you’re trying to say here? 

Mat Cashman

It’s a good intuitive way to think about it, but that’s not necessarily the case. That’s one of the most important parts of this discussion. The ratio by itself is not a trading signal; it’s information. If implied volatility is running over one in that situation, the ratio is running over one. In that case, it means the implied volatility is essentially 135% of the comparable historical volatility. 

Then your next question should always be, “Okay, if that’s the case, why? Why is that the case?” Maybe earnings are coming. Maybe there’s another unknown event that’s on the horizon ahead. Maybe the underlying has been unusually quiet during that 30-day historical period, and the options market is just reflecting an expectation going forward that quiet period won’t continue. Or maybe there’s a different level of uncertainty being incorporated into option prices broadly across the board in multiple different asset classes. So the relationship gives you something that’s worth looking at. It gives you a metric to start to dig in. It doesn’t automatically tell you what to do, and it doesn’t tell you what conclusion to reach here. That’s one reason why I’m always careful when people ask me about words like “cheap” or “expensive” when we’re talking about implied vol. That makes it sound like the analysis is already done. When what we’re really talking about, and the point we’re at here creating this ratio, is really where we’re starting the analysis, not the endpoint of the analysis. And I think that’s important. 

Jeff Praissman

In this day and age, there’s no shortage of data points. You had mentioned, though, comparing similar periods. So why is that important for people looking to see whether or not options may or may not be overpriced? 

Mat Cashman

Yeah, that’s really interesting because historical vol isn’t one fixed number either, right? If you’re looking at a 30-day historical vol, what you’re really looking at is the last 30 days of movement combined into one data set. So, for instance, a stock could have a 10-day historical vol of 50 because it just experienced a big move in the last 10 days, but its 90-day historical vol could still be 25 because you take those 10 days that are a 50 vol, and then you add the other 80 days that are well below 50, and it averages out to a 25 vol over 90 days. Both numbers, interestingly, are correct, right? The 10-day historical vol could be a 50, and the 90-day historical vol might be a 25, but you’re gonna look at it and be confused by it sometimes. They’re just measuring different windows of time, different windows of realized movement. And the same thing applies on the implied volatility side as well, which can be confusing. The 10-day option and the 90-day option are incorporating different expectations across very different periods of time. 

And so if we’re gonna compare implied and historical vols, we need to be thoughtful about what periods we’re comparing. Otherwise, you can produce a very precise-looking ratio that really isn’t telling you anything about what you think it’s telling you. 

Because if you’re comparing a 10-day historical to a 90-day implied, you’re looking at two very different things. And keep in mind, I always think about this in terms of historical vol looking backward and implied vol looking forward. You’re already looking in two different directions, right? The relationship is the best that we can get as far as those two things are concerned, but make sure that you’re trying to match it up as close as you possibly can. That’s what I would try to say. 

Jeff Praissman

Yeah, because comparing a 10-day and a 90-day is really comparing apples to oranges. 

Mat Cashman

Agreed. Yeah, exactly. 

Jeff Praissman

But that does point out something that even a lot of traders, a lot of people say all the time. They’ll say “the implied vol,” “the IV,” as if there’s only one number for that stock, but there really isn’t. 

Mat Cashman

Exactly. And oftentimes when you look at a user interface for something, it’ll show you the implied volatility for that month’s options. What it’s really showing you is the implied volatility of that month’s at-the-money straddle, essentially, right? But there really isn’t one implied volatility number associated with an underlying. Every option has its own implied volatility that’s associated with the price of that option that’s trading at that moment in time. And thus, all the options have their own implied volatility, but different expirations have different implied volatilities as well. And then different strikes have different implied volatilities. And that gives us what we call a volatility surface rather than one single number that you pluck out of the ether. And so the near-term options, in that case, might reflect something significantly different as far as volatility because there might be something like an earnings announcement that falls inside that expiration that doesn’t fall inside the longer-dated options, right? The longer-dated options might show you something totally different. And within the same expiration, out-of-the-money puts, at-the-money options, and out-of-the-money calls can have different implied volatility levels too because of the strikes that are associated with them. 

So when someone tells me, “Okay, well, the implied vol’s 40,” I still have a lot of questions left that I ask, which is, “Okay, well, which expiration are we talking about? What strike are we talking about? What part of the surface are we actually talking about as to where that 40 vol is coming from?” Because it can be so many different things. And so it always, in my mind, comes back to context. Build a context around that number because the number itself doesn’t tell you anything unless you have reference points to understand what you’re talking about. 

Jeff Praissman

Yeah, and I think one challenge that people face with implied volatility is just saying it’s 20%—it’s an annualized percentage. It doesn’t necessarily feel that intuitive what it actually means for the pricing. So how can you make that more tangible? 

Mat Cashman

That’s a great idea, and I talk about the Rule of 16 all the time when I talk about implied volatility, or a lot of the time. And the Rule of 16 can be really helpful to try to take that annualized number because, remember, implied volatility is an annual metric. 

And so when people ask me, “All right, well, what does the 20 vol mean? It’s an annual number. I trade my options on a minute-by-minute or a daily basis. What does the annual number have to do with anything?” Well, the Rule of 16 can be really helpful in that situation. Suppose options on a $100 stock are trading at 20 vol, a 20 implied vol. And so as a rough framework, which is basically what the Rule of 16 is—remember, the Rule of 16 came from a bunch of traders standing in a pit with cards in their pocket, writing this down on the back of an envelope. So it’s not complicated. It was designed in a way that makes it really easy to use and quick. And so as a rough framework in that situation, remember: $100 stock, 20 vol. You take the 20 vol and you reduce it to a decimal. So a 20 vol is 0.20. You multiply it by the stock price and you divide by about 16. 

What that does is it gives you a price expectation for movement in the underlying on any given day. And so instead of only saying these options are trading at a 20 vol, you can translate that into something more like, “Okay, at this 20 vol level, the options are reflecting future movement on average of about a buck and a quarter in the stock on any given day.” That’s basically like one standard deviation of movement for that stock based on the implied vol levels of the option. So now I have something I can kind of visualize, right? If you’re walking in and the vol that you were trading was a 20 vol, and you know that’s basically forecasting a buck and a quarter of movement going forward, and the underlying suddenly moves six or seven bucks in one day, now you have a framework to be able to look at that big move and understand how big it actually was relative to the amount of movement that had previously been reflected in the option prices, which is what they’re trying to forecast going forward, right? And so that’s a big difference in being able to look at a six-dollar move and say, “Oh, that’s like 3.75 standard deviations,” instead of just, “Oh, the thing just moved six bucks. What do I do now?” Right? 

Jeff Praissman

And interesting enough, too, right? After a big move like that, or any move, or any non-move, the implied volatility of options really can change significantly, right? It’s constantly changing. It’s a dynamic feature. It’s not static where it’s gonna stay that same 20% forever. 

Mat Cashman

Yeah. The implied volatility is tied to the price. So if the prices of the options are changing, it’s most likely that the implied volatilities are changing too. And that’s exactly what you just said. That makes this a very dynamic relationship. 

Suppose that underlying makes a huge move, and the options that have been trading at 20 vol are now adjusting to 40 vol. A lot of things are happening behind the scenes there as far as the vols are concerned, not only with the stock moving, but how the options are priced. The underlying price has changed, the implied volatility has changed in that situation, and the realized volatility picture may be changing as well. The implied volatility is moving, and therefore the relationship between the implied and the historical vols may be changing there. So the useful question becomes, when you’re looking at something that moves like that, what does this new implied volatility level tell us about the environment the options market is pricing now? 

Maybe now that this has happened, it’s pricing something completely different going forward. And then, again, you’re not just simply looking at that new, potentially 40 vol and saying, “Okay, is 40 high now?” Right? Again, when you look at that new implied vol, compared to what? 40 vol is high compared to what? Now, do I have a different set of expectations for the price of the underlying going forward? And do those expectations match the 40 vol that I’m now seeing in front of me that used to be 20, right? That’s a big difference. And I think about it as a completely different framework to think about the vol that you’re looking at on the screen. And so that’s why I find it useful. 

Jeff Praissman

So, instead of just looking at the relationship on a single day or today, would you track it over time then? 

Mat Cashman

Absolutely. You can track that relationship over time. You can track the ratio over time. And I think that’s where this framework gets even more useful. Suppose the implied-to-historical ratio for an underlying has generally been around something like 1.05. Remember, earlier we showed those two implied and historical ratios spitting out 1.35, meaning the implied vol was 135% of the delivered vol. But in this case, let’s say historically the ratio of this thing is 1.05. And now it’s suddenly 1.50, right? It goes from 1.05 to 1.50. So 105% of delivered vol to 150% of delivered vol. Now, something has changed there, but that still isn’t the end of this analysis. You have to ask, why did the ratio change? Did implied volatility—did your numerator go up? That’s one way that ratio can change from 1.05 to 1.50. 

Did historical volatility fall, right? Did the denominator drop? Because that’s the other way that ratio can move from 1.05 to 1.50. Or did both happen at the same time? Because the ratio’s made up of two moving variables, right? The implied vol and the delivered vol. And so the ratio tells you something about the relationship, but you still need to look underneath it to understand which part of that is moving to make the ratio move. And that’s an important step in this analysis as well. 

Jeff Praissman

So, can this framework that you just described be used to compare different underlyings? 

Mat Cashman

Yeah. I think that’s kind of the interesting part about this framework, is that it reduces all that stuff down to something that you can actually look at, even cross-asset sometimes. And that’s one of the things I really like about it. Absolute volatility levels can make cross-asset or cross-underlying comparisons really difficult because of that situation that I talked about earlier. 

If the options on one stock normally trade on a 20 vol and another stock’s options normally trade on a 70 vol, simply saying, “Okay, the 20 vol’s a buy and the 70 vol is a sale,” that’s not gonna get you very much. And it might work, but it also might not. 

Because if you compare each implied volatility level with its own set of relevant historical vol, or with its own normal historical relationship, you have a much more standardized framework that you can use to potentially compare two cross-assets that might not necessarily be comparable in any other way. You’ve removed some of the absolute level bias, which is what I’m really looking for when I do this. That doesn’t mean the two underlyings suddenly have identical risk. They obviously don’t. It simply gives you another way to evaluate the relationship between what each underlying has been doing and what its options are actually currently reflecting. 

And that, I think, is a more nuanced way to look at those relationships. 

Jeff Praissman

And Mat, you keep coming back to the word “framework” in your description. Is that on purpose? 

Mat Cashman

I use that word a lot because I think it’s a useful way to approach these metrics. Options and the metrics that options give off, right? Like you said before, there’s a lot of noise out there. There’s so many data points that you could look at. And options particularly give us a tremendous amount of information, but it can be very overwhelming, right? Every strike has a different vol. Every month has a different vol. And so very few individual metrics actually give us any kind of complete answer. Delta doesn’t give you an entire picture. Historical volatility doesn’t give you a complete picture. And guess what? Implied volatility doesn’t give you the complete picture either. And so the implied-to-historical volatility ratio helps, but it doesn’t give you the complete picture either. They’re tools, right, for organizing information and essentially asking better questions. If a relationship looks out of whack or if it looks unusual, now I know where I might wanna start looking more closely. And that’s really valuable because of what you said, right? We get so much data that it can be very overwhelming and kind of disorienting in a way. So I don’t need this metric to make the decision for me, but it helps to be able to perk my ears up when I see something like this move in an unusual way. 

It’s a great way to have an idea of where to look. And then you start digging into, “Okay, why did this thing move in this way?” And that gives you a lot more information about what’s going on. 

Jeff Praissman

So Mat, this has been great, and just kind of wrapping this all up into sort of a takeaway. 

If someone’s looking at implied volatility on their option chain, on their trading platform, hopefully Trader Workstation or Desktop, what is the takeaway that they should have? 

Mat Cashman

Yeah, so I would say the real takeaway here is don’t just stop at the number, right? Don’t just look at it and say it’s a 20 vol and that’s the end of the story. The implied volatility number is really the starting point of where your analysis begins. It’s not the end; it’s the beginning. And then start asking those questions. Compared to what? That’s what I really want people to take away from this. It’s a 20 vol. Is that high? Is that low? I’m not sure. Compared to what? Because compared to the underlying historical movement can be one thing, or comparing the 20 vol to the next expiration, which might be a 35 vol or something. A 20 vol on one strike compared to another strike, that can be something. Or compared to another underlying, or even compared to its own history of where the implied vol is. Because once you give implied volatility some context, right, if you build this framework around it, the number starts becoming much more informative. You’re not just looking at an annualized percentage anymore, which is what that is. 

You’re starting to understand what the options market is actually reflecting about future expectations and future uncertainty. And potentially, if you dig in a little bit more, you might understand why some of those expectations are what they are based on the price and the commensurate implied volatility level. 

Jeff Praissman

Mat, this has been great, as always. Love having you come in here and talking options. And again, for our listeners, you can find more from Mat at optionseducation.org or occ.com, as well as on our website, interactivebrokers.com. Go to Education, click on webinars, podcasts, and articles. Thanks again, Mat, for stopping by. 

Mat Cashman

Thanks for having me.