Automated Options Trading #2: What is Expected Value (EV) & why should you care?
Вставка
- Опубліковано 4 сер 2023
- optionalpha.com - Expected Value (EV) is a fundamental concept in probability theory that provides insight into the outcome or value of a random variable. To put it simply, EV is a weighted average of all possible outcomes. It serves as a powerful tool for decision-making and risk assessment. In options trading, calculating the EV of a trade at expiration is useful for evaluating its potential profitability. To accurately calculate the expected value for an options trade, we found it critical to assess the probabilities and associated payoffs not just for a single profit/loss scenario but for every profit/loss scenario.
Learn more about how we calculate EV here: optionalpha.com/help/understa...
Learn more about how we calculate probabilities here: optionalpha.com/help/calculat...
Want to follow this series? Click here to Subscribe to updates: ua-cam.com/users/OptionAlpha?s... - Навчання та стиль
I've been waiting YEARS for someone smarter than me to properly apply the EV equation to vertical spreads. This is huge for me.
Your detailed explanation clarified EV greatly. This is taking BOT performance to a whole new level! Thanks Kirk!
Very welcome!
Terrific training EV video. Factoring in partials vs only binary outcomes is critical. Thanks
Great video. I look forward to discussing the nuances of EV in the Option Alpha Community.
Hi Kirk
Been following you for many years now and so glad to see you are now bringing Expected outcome (Expect value) into play now. I believe this is critical to successful trading.
There are very few tools available that provide this powerful information...
Loving the platform and the progress.
I managed to cobble together a loose approximation of this on an Excel spreadsheet. I did have to take a few shortcuts. For example, the break-even point is rarely on a listed strike, so I had to do a _linear interpolation_ of the _PoP%_ between the short and long strikes, and also ass-u-me that the probabilities between the two (actual) strikes and the (virtual) breakeven _”strike”_ were *LINEAR.*
Still, it beats taking the _Probability of Profit_ and _Return on Risk_ figures from my platform _(ToS)_ and adding them together as a rough sort of _”figure of merit”._
This is what I want to learn. You are doing the right thing here.
As you know, a lot of money can be made by closing off trades a little early. I believe it we have an EV, that better our net profits.
Keep it coming.
Thank you for posting this, Kurt. It is very helpful.
I have been trading multi-leg option strategies for about 5 months, and I have been wondering how to know whether IV is high enough to support short premium strategies. I believe EV, as you describe here, is the answer.
I felt like I couldn’t lose with vertical spreads and ICs in March and April, but I took a beating in new ICs through June and July following the implosion of IV and the AI frenzy sparked by NVDA’s May earnings. I switched to lower-IV strategies in July with good success, but I believe I would have switched sooner, to my benefit, had I been considering EV.
Excellent presentation! This explanation really helps clarify some ambiguities I've had in deciding on which trades to make.
Glad it was helpful!
Great video & kudos once again to OA for releasing cutting edge research and features.I think Nassim Taleb first suggested in Fooled by Randomness that, I paraphrase, option sellers eat like chickens and poop like elephants. It seems to me that EV holds the key to avoiding this trap which I think is the result of managing each trade separately rather than seeing them as just another iteration of a model that should (ideally) provide a little bit of edge over time. Allowing that edge to play out is going to require the absorbtion of lots of losses. Trading like the casino sounds easy but it is not, let the bots (and/or your rules) be your dealers and stop looking at individual trades, it's probably the hardest part of trading because we just don't like to lose.
Excelent video 👏
I have been doing some EV calculations by hand, and it appears the first-order approximation (ignoring the probabilities along the sloping P/L line) works well as long as the residual probabilities between max loss and max profit are small. For example, in your RTX example 13:20, the first-order EV is $8.94 compared to your higher-fidelity calculation of $6.61. I wouldn’t say this difference is huge relative to the values of max profit $29 and max loss $71, but the residual probabilities (100% - probability of max loss - probability of max profit) only come to 4% in that example.
I did another EV approximation by defining profit and loss triangles between the strikes in the P/L diagram, and allocating the residual probabilities (4% in the RTX example) to max loss and max gain according to area-weighting of the corresponding profit and loss triangles to the sum of the areas of the two triangles, which gave me an EV of $6.68, near your higher fidelity calculation of $6.61.
Excellent technical content
Thanks!
Thanks for for the video. I was following the thread in the Option Alpha community that I think may have prompted this explanation.
My understanding is that all probabililities are calculated to expiration, correct?
7:36 The language used "stays above" and "starts trading below" seems to be a misstatement. I believe this should instead say "closes above/below at expiration". Other platforms do calculate "probability of touch" as the (higher) probility of breaching a strike during the trade, but I don't think this is the calcualtion that OA is using.
10:07 This statement seems to imply "all outcomes" are considered, but in reality only "all outcomes at expiration" are considered. This can lead to some misunderstanding amongst less experienced traders, who do not understand how different the risk graph looks when plotted at various intervals during the trade compared to the sharp angles at expiration.
-Keith
Correct these are "at expiration" and I'll try to throw that verbiage into future videos. The payoff diagrams are always "at expiration" by default.
Thanks Kirk, looking forward to seeing how this Bot performs. Is it available to clone?
So people have created cloned versions in the Community but my plan was to share the template once I do more testing, if it needs adjustments or changes.
EV = area of triangle above 0, minus area of triangle below zero?
Than you for this Video. To the best of my knowledge you did not show in this video how to calculate the Probabilities associated with the values between the Max. Profit and Max. Loss. If the break even point falls exactly in the middle of the two strikes so the Value of the Partial profit and loss is Zero so it doesn't matter. But if the Break even Point is not exactly at the mid point of the Two strikes - how do you calculate the Probabilities for all the partial profit and losses to add to your Max Profit and Loss Formula. Thank You
Great question and we cover how we calculate it here in this article: optionalpha.com/blog/how-to-calculate-expected-value
TL;DR version - we calculate the partial profits and partial losses of all possible outcomes between strikes, and their respective probabilities of occurring.
I do not have high confidence in this system in low low IV .... in high IV pretty sure it would work ... hmm try to combine some filters from your High IV IC one .. have broke the wing down in that with very good results ... what an adventure .. and like Kirk says some times the best trade is no trade .. for sure we need to get compensated as sellers high IV will do this .. and many times if you look at diff DTE you get compensated better .. no rhyme nor reason that I can put my finger on lol but the 90 DTE and higher are my type of trade " )
thinking a bit on this .. I am sure you know FV " fair Value" normally when market up FV is below market and down FV is above market value .. there are some others come to mind Theo Price .. but I would not be sure any substitute indicators or metrics .. that could be built out in a bot .. but EV I think lives kind of sort of with FV and Theo Price .. and for sure would be good checks before and entry I have a test one running solely on Alpha Rank ... I buggered bot up . by switching to just SPY .. but it was doing pretty well with ye old Kirk uncorrelated ones LOL but a proper check .. should be just one underlying ???
they're day trades? but how do we get around PDT? without 25k. I think holding over night gives us added risk. but if same day is red still next day things might change. It's almost like coin toss
These aren't day trades unless you do a 0DTE or 1DTE trade (timing dependent). The testing I'm doing for these are currently on 5-10 days out so these would all avoid any PDT rules.
there's still an issue with the calculation as they do not match either TD or Schwab in terms of POP/PMP. I dont know which is correct but it appears OA's calculation is much more optimistic. any feedback on this?
Those companies don't publish how they calculate their POP/PMP but we do here: optionalpha.com/help/calculating-probability