How To Trade Options Efficiently: Part 1
Updated: 2020-05-23.
An option is a financial instrument. Like stocks and bonds, options can be bought and sold with the intent to profit from price movements. While options are designed to provide optionality they also have the potential to generate a stream of income. It goes without saying they have the potential to lose money, too.
A typical trading strategy will aim to achieve:
- High Return – generating the most income per trade
- Low Volatility – generating similar amounts of money per trade
- Repeatable – consistently repeating the two goals above
Let’s look at brokerage selection, opening positions and managing positions to see if there’s an optimal approach to achieving these goals.
If you search “options trading strategies” you’ll find no shortage of content, both free and paid. Some websites offer sensationalized “secrets” and promises of returns. Others offer matter-of-fact details that explain various trading tools and how each one works.
The problem with sensationalized claims and tool explanations is that they don’t explain how each claim or tool performs. There is no data. I’m not willing to risk my hard-earned money in something that isn’t evidence based and consistently repeatable.
Without transparent data available on the internet I was forced to do my own research. I decided to publish my findings and use the insights gained to outline a systematic, repeatable, quantifiable approach to options trading.
Contents
Basics
This guide assumes a working knowledge of options and their mechanics. Let’s review the high-level concepts in a brief refresher.
Stocks
When buying or selling a stock there is a single dimension: price. After placing a stock trade, the price will either go up or down based on unpredictable market dynamics. There is a 50/50 chance of making money.
Options
When buying or selling an option three additional dimensions are introduced: strike price, time and volatility. Each dimension can influence the price of the option and consequently the probability of making money. Options have probabilities ranging from 99.9% likely to make money to 99.9% likely to lose money. More on this in a little bit.
Strike Price
The strike price is the price at which an option contract can be exercised. The relationship between strike price and the price of the underlying stock upon which the contract is based determines an option’s intrinsic value.
Time
Options are expiring assets with a defined lifetime ranging from less than a day to nearly three years. The value of the time remaining until an option expires determines an option’s extrinsic value. The farther away an option’s expiration date, the more extrinsic value an option will have. As time passes and an option’s expiration date nears, an option’s extrinsic value approaches zero at an exponentially accelerating rate. This rate of decay is represented by theta.
Together, strike price and time, or intrinsic and extrinsic value, determine an option’s price which is called premium.
Volatility
Volatility determines the price per unit of time and uses probabilities to estimate how far an underlying’s price is expected to move over the next 52 weeks. The Black Scholes model is an academic formula the describes how the time value and probabilities are calculated. Modern brokerages use other models such as splines, to implement interpolation for determining value.
Trading Fees
Like buying or selling a property or stock, each purchase, sale, exercise, or assignment of an option may incur a transaction cost. For options they are the: base trade fee, per-contract commission and exercise and assignment fee. In addition to trading fees, some brokerages charge monthly subscription fees for access to real-time options market data.
Brokerage Selection
A brokerage account is the starting point for placing trades. We’ll want to utilize a brokerage with the lowest trading fees. For the purposes of this ranking we will be looking at the:
- cost associated to open and close a single stock option contract with a premium over $0.10
- fee associated with option exercise and assignment
- market-data subscription fee, if any
The top three brokerages for options trading based on fees are (these are NOT affiliate links):
#1 Robinhood – FREE
Robinhood has $0 base trade fee to open or close a position, $0 per-contract commission and $0 exercise and assignment fee. Real-time market data is also $0. They also offer $0 trade fees on equities, too. Both equity and option trades are executed in real time.
Is there a catch? No. However, if you intend to use Robinhood for purposes other than trading options, such as for margin loans or housing your retirement accounts, there are better alternatives.
#2 tastyworks – $1.20
tastyworks has a $1 base trade fee to open a position and $0 fee to close a position, $0.10 per-contract commission to open and close a position, and $5 exercise and assignment fee. Real-time market data is FREE.
#3 Interactive Brokers – $2.00
Interactive Brokers has a $0 base trade fee to open or close a position, $0.70 per-contract commission to open and close a position with a $1/trade minimum, and $0 exercise and assignment fee (the stock trade fee minimum of $1 is assessed). Real-time options market data is $1.50/month and is waived if monthly trade commissions total $20 or more. Interactive Brokers is known for having the lowest margin rates in the industry.
Winner: Robinhood
Opening Positions
When trading an option we have to pick an underlying, determine whether we want to buy or sell, select contract duration, identify what delta we want to trade at, evaluate the volatility of the underlying, and finally choose an instrument. Phew! Let’s look at each one and see if there’s an optimal approach.
Underlying
When choosing an option underlying there are two key attributes to consider:
Liquidity
Liquidity is hands down the most important attribute of an option. It’s what allows a trader to get in and out of a position at a fair price. Liquid options also have more accurate probabilities, per the Black Scholes model, than illiquid options. Option liquidity is typically measured by volume and open interest. Highly-liquid options will usually have a tight bid-ask spread, often times as narrow as a penny or two.
Options on SPY are by far the most liquid options on the market, having a volume multiples higher than the next most liquid underlying. Barchart.com freely lists the volume of options trades for various underlying.
Liquidity winner: SPY
Risk
The total risk of an investment is comprised of two elements: systematic risk, sometimes also referred to as market or undiversifiable risk, and unsystematic risk, sometimes also referred to as specific, diversifiable, residual, or idiosyncratic risk.
Systematic risk is risk the entire market will experience – interest rate changes, recession, natural disasters, political unrest etc.
Unsystematic risk is risk specific to an individual stock, industry or sector – a company cooking its books (Enron), airline industry scares (Boeing 737 MAX) and tech and housing sector bubbles. Unsystematic risk can be mitigated through broad diversification, effectively lowering the total risk of the trade.
Risk winner: any broad market ETF such as SPY, and to a lesser extent, sector-specific ETFs such as EEM and XLF.
Winner: SPY
Action
There are two possible actions: buy an option or sell an option.
An option seller can be profitable in 3 scenarios – when the underlying price:
- stays the same
- moves favorably
- moves slightly unfavorably
An option buyer can be profitable in 1 scenario – when the underlying price:
- moves favorably fast enough to overcome theta
The option seller has 3x the opportunities to be successful than the option buyer and isn’t racing against the clock.
Winner: Sell
Duration
When we sell options we are selling assets that depreciate in value as time passes. We can sell an option for $100 today then buy it back at a later date for, say, $75 making $25 on the trade. This depreciation factor allows us to sit back and profit from the passage of time.
As an option’s expiration date approaches it loses value faster. Great! Let’s sell the shortest duration options we can find so we can make money as fast as possible.
Not so fast!
As an option approaches expiration, its price becomes more volatile and behaves less like the stable unwinding of a clock. That’s not what we want. In fact, much of the potential profit (theta) that’s available near option expiration is lost due to this volatility (gamma).
Let’s look at the performance of selling 45 DTE strangles and closing the trade early at various increments.
By closing the positions at 21 DTE, 21 days before they expire and avoiding the fastest decay time period, profits are actually higher than participating in seemingly most profitable time to hold an option.
Ok, so should we instead aim for the longest duration option to have a slow and steady profit? That’s not quite ideal either.
Let’s hypothetically sell 16 delta strangles on SPY on the first trading day of every month for 10 years, starting in 2005. We’ll do this with 45 DTE, 75 DTE, and 110 DTE positions.
It turns out the 45 DTE options had the greatest P/L.
To identify the optimal duration we want to find the point on the option premium time curve where additional time results in a diminishing return of premium received.
This happens to be around 45 DTE for a 16 delta put on the SPY. Selling options of this duration allows us to partake in the sharpest declines in option price while exhibiting the greatest daily P/L.
Winner: closest to 45 DTE
Delta
Delta is a number that says a lot about how an option behaves. For our purposes we will be looking at its usefulness in determining the probability of having a winning trade.
To calculate the probability of profit (POP) or win rate when selling options:
- ignore the +/- sign in front of the delta value
- subtract the delta value from 100
An option with a delta of 50 means there is a 50% chance the trade will be profitable when held to expiration. A delta of 5 means there is a 95% chance the trade will be profitable when held to expiration. A delta of 90? Only a 10% POP.
Over the span of ~3100 trades the realized win rates tend to align with expectations.
As a general rule, when win rate increases the premium collected when selling an option goes down. Less risk, less reward.
Winner: dependent on trader’s risk and return targets.
Volatility
Volatility is rooted in the concepts of probability. Remember the standard (bell) curve from high-school math class? No? You were sleeping or skipping? Here’s your SparkNotes:
Options have an attribute called implied volatility or simply IV. IV is a percentage value that describes the expected annualized price range of a one-standard-deviation move of an underlying. That’s a mouth full. Let’s make this easier with an example and picture.
Suppose we have a stock, XYZ, that’s worth $100 and it has an IV of 25%. This means there is a ~68% chance (one standard deviation) XYZ’s price will be between $75 and $125 12 months from now. Depicted visually, it looks like this:
Why do we care about volatility? When IV increases, the price per unit of time increases. This means we can:
- receive greater compensation for trading at the same delta (POP)
- trade at a lower delta (i.e. higher POP) and receive the same amount of premium
Examples of strategies that are based on volatility include trading when IV is above / below a target and trading continuously without regard for IV.
Winner: dependent on trader’s risk and return targets
Instrument
There are two instrument choices when it comes to options: call and put. These instruments can be used on their own or they can be used as building blocks for more complex strategies such as a strangle or straddle.
Let’s evaluate the basic instruments. More complex instruments provide additional “features” such as defined risk but come at a P/L cost and incur additional frictions.
When selling puts, profits are made when the market rises, stays the same, or falls slightly. Selling puts tends to be profitable in bull markets.
When selling calls, profits are made when the market falls, stays the same, or rises slightly. Selling calls tends to be profitable in bear markets.
Since we do not know the kind of market we’re in until after it occurs it is not possible to optimize instrument selection at the time of order entry. However, knowing how each instrument performs in all markets helps us select an instrument for “all purpose” use.
Puts are consistently profitable over time no matter which strategy is implemented. Meanwhile, selling calls is essentially unprofitable unless leverage is used and even then it’s still a potentially-unprofitable strategy.
Winner: PUT
Managing Positions
Managing positions is all about proactively and mechanically acting on trades to improve win rate and P/L. As we’ll soon see, the Wall Street proverb “cut your losses but let your profits run” isn’t always the best strategy.
Profit Targets
The concept of profit targets is simple: take profits when trades are profitable. No one ever went broke taking profits.
A typical strategy for managing positions is the “stop loss.” When a position falls below a certain P/L value, close the position as a loss and move on to the next trade. Managing winners does the exact opposite. When a position crosses a specific profit target, close the position as a winner and move on to the next trade.
Think about it for a second – markets ebb and flow all the time. Using a stop loss ensures we exit during a market lull while using a profit taker ensures we exit during a high point.
Of course, this concept means nothing without data to support it. Let’s take a second look at the chart from the instrument section above.
The first 3 rows show us the performance of taking profits early – at 25% of max, 50% and 75%. When compared to using various stop losses early management improves total P/L across the board.
Early management increases the win rate vs using a stop loss more often than not.
Managing trades early improves risk-adjusted returns vs using a stop loss more often than not.
Winner: use profit targets
Early Exit
Managing winners has the benefit of closing trades earlier, avoiding the most volatile time for an option. We want to spend most of our time in the boxed area, letting the option calmly depreciate in value so we can buy it back later for less.
Let’s take a second look at the P/L chart from the duration section earlier in this guide.
Using the 21 DTE and 7 DTE strategies, it was consistently more profitable to both set a profit target and close the trade before expiration than to use a profit target alone.
This means some positions will be closed without them ever reaching profit targets. In fact, they may even be closed at a loss.
Winner: close trades before expiration.
Summary
Setting up a trade requires making many decisions. The following configuration appears to lend itself to the greatest likelihood of success:
- Brokerage: Robinhood
- Underlying: SPY
- Action: Sell
- Duration: closest to 45 DTE
- Delta: dependent on trader’s risk and return targets
- Volatility: dependent on trader’s risk and return targets
- Instrument: Put
There are systematic, data-driven mechanics that can be applied to open positions to improve trade profitability and consistency of profits:
- Close trades at profit targets
- Close trades early if profit target is not reached
- Manage risk at order entry by selecting delta and number of contracts, as opposed to after the fact using stop loss orders
A key tenant to success is eliminating the emotional side of trading. Staying mechanical will help ensure the probabilities play out.
Check out part 2 which covers leverage, capital allocation, overstatement of delta premium, overstatement of duration premium and tax-efficient scaling. Part 3 covers trade occurrences, capital efficiency and the Micro e-Mini Futures product.
March 30, 2019 @ 7:26 pm
I think it is in very bad taste to use so much of Tastyworks’s research and charts and then recommend a different brokerage.
At the very least you should more clearly credit Tastyworks since the vast abundance of your offering is taken directly form there excellent work. At TWs very low commissions do you really think it is fair to suggest folks go elsewhere? No pun intended but I think that is all very bad taste.
~ A tasty national since 2012
March 30, 2019 @ 8:07 pm
I understand where you’re coming from and appreciate the feedback, Joe. I’m all for attribution. The intro clearly mentions TT and speaks to their ongoing research, that this is largely a summary of their work and each image is linked back to the respective source/study. What additional attributions do you suggest?
As for the platform, the facts suggest Robinhood is the optimal choice, which is both this post’s first topic (brokerage selection) and objective: identify the optimal approach to trading options. It would be a disservice to the readers and myself to suggest otherwise. If this was a sponsored post, then fully agree.
April 11, 2019 @ 2:29 pm
I have a number of questions and am currently giving this a try in a test account in IB. First, I noticed in your sample you didn’t wait for 50% profit on any of your transactions. They range from 6% to 32% and look to average about 18%. Is there a reason for the early closing? Next, you mention that one should only invest about 20-30% of capital in this approach. Is the 20-30% based on the amount of the underlying Stock assuming all open options are exercised or an amount one is willing to loose whether that be selling for a loss at 4x, 5x, or whatever number is reasonable to the investor prior to the 21 day sell if desired profit was not met mark? Is there a reason for the laddering having multiple trades on different Sell dates for the same expire date with essentially the same SD? If the 45 DTE is the optimal number then why wouldn’t you just trade the full number of options on the Tuesday 6.5 weeks prior to the Friday Expire date? Lastly, you recommend to use Robinhood, however you use IB for your personal trading, is there a reason for this? Thanks in advance.
April 11, 2019 @ 10:59 pm
Hi Mike! Happy to hear you’re diving in with a test account. Great questions, too!
Answer 1: Yes! By managing early (even earlier than 50%) several things are accomplished.
1a) It improves the win rate. These closed trades will never “go bad.”
1b) It can improve return on capital. One of the two green columns in the trade log is APY, which is how I measure my capital efficiency. Even though a trade may be held for only a day or three, making 10-20% in that time represents a significant return. Using the 5 delta puts near the end of March as an example, holding those till expiration would have yielded the equivalent of 1.4% annually. By closing early I’m getting the equivalent of 5.5-8.4% (actually more, but my date math doesn’t include time stamps so I round up to the next day, making APY artificially lower).
1c) It can reduce risk. One aspect of managing early is that capital can immediately be redeployed on another trade. Personally, I prefer to take downside risk off the table and wait until the next day (or longer) to place another trade. Should I take profits in the AM then the market turns down in the afternoon, I’ll have avoided not only the loss of accumulated profits, but I’ll then be able to buy into the elevated VIX. Higher VIX = higher value per unit of time remaining till expiration.
1d) Consider 50% max profit the ceiling. Trades can be managed at 50% or they can be managed [even] earlier.
Answer 2: this is 20-30% of notional value. For example, if I sell a single put option with a strike of 260, that represents $26,000 of notional value ($260 underlying * 100 shares) – it’s also the theoretical max loss. If I have a 100k account, that single trade will represent 26% of my available capital. Granted, this mechanic is designed to protect from the worst case scenario [since the Great Depression]. Assuming the market never drops that much ever again, these numbers would be too low. Inversely, if it drops lower, these numbers are too high. If we were to use 1.5SD or 2SD options, I suspect the numbers would allow greater allocation. I unfortunately don’t yet have any data on this to validate.
This of course begs the question: how does one get started? The solution is to use put spreads (eg: sell a 260 put and buy a 250 put to mitigate the downside risk). The premium collected will be less, but so will max loss / capital allocation (eg: 26k exposure from put sold – 25k protection from put purchased = 1k of capital at risk). One thing to keep in mind: the 20-30% allocation is based on 1SD strangles.
Answer 3: That exact question is addressed in part 3, which I’m writing now :). Short answer: it’s about placing multiple individual bets vs one large bet in order to realize the statistics. Yes, I have a 10 lot strangle in my trade log. I need to increase my discipline :). I have capital to allocate hence the daily trades. The deltas are around the same but, the VIX at the time of order entry, and consequently the premium collected and strikes, was different. Because I’m trading such low deltas the variance in these attributes is nominal. However, if I was trading closer to the money there would be more material differences.
Answer 4: I went with IB as a means to obtain cheap borrowing rates. I have no intention to borrow but I like having the ability should a need/opportunity arise. My deferred capital gains makes borrowing and paying low simple interest an attractive workaround to paying 15% in taxes. I opened the account using an affiliate link – it offered a lifetime 0.5% reduction on the first 100k of margin. In other words, the 0-100k tier is priced the same as the 100k-1M tier. The other reason is I tend to experiment with things – futures, futures options, and even considered carry trades using dividend stocks to offset the margin interest. Also, the paper trading feature and no outbound ACATS fees is nice.
April 12, 2019 @ 7:48 pm
Thank you for the response and the additional detail.
1. That is what I had thought, just wanted to confirm. Thanks
2. Follow-up – As I see it there are 3 possible scenarios for losing money (i.e.. Risks for capital loss). Let’s assume we are trading a single put for SPY at 1SD let’s say $278 for $1.50
a. Prior to the [21 day close no matter the situation] the underlying position transitions into-the-money to say $260 and is exercised. This would require purchasing 100 shares at the strike price or $27,800 so Capital at risk is $27,800. The current holding could be sold at a loss or calls could be sold to see if the holdings could be exercised away with no loss but no guarantees here. The sure way to mitigate this situation is to trade options which cannot be exercised early (i.e.. European-style options). SPX Index Options are traded European-Style therefor they cannot be exercised early. I would think this would be a big plus for SPX Index options over SPY.
b. Prior to the [21 day close no matter the situation] the underlying position tanks. Let’s say it goes down to $100 and is exercised. Again, this would require purchasing 100 shares at the strike price or $27,800 for a position that is now worth only $10,000. Options are really limited here. You could sell and take a $17,800 hit or hold on and hope it recovers over time (Hope you’re not using margin 😊 for your trading). You can mitigate this risk as you suggest in your prior reply by setting up a Put spread and buying a put at say $250 for $.30. This would limit your initial Put profit by $.30 ($1.50 – $.30 = $1.20) but would mean if this event occurred you would only be out $2,800 which is better than $17,800 and the capital risk is only the $2,800 max loss. Again, there may be some additional positives here trading SPX Index options although remember everything is multiplied by a factor of 10 for SPX Index Options. Instead of utilizing a Put Spread I would sell the Put as normal at 1SD and add an alarm at the 5x level for the Put. The option will not be exercised but it does not mean it cannot be closed. If the option increased to the alarm setting the alarm would sound and one would have the discretion to either close out the position for a loss or hold on and take on additional risk. In a 10x example using SPX the initial Option would be $15 and an alarm would be set for $75. If the alarm triggered and discretion was used to close the option a loss of $6,000 would be incurred. You will notice that using SPX Index options this way the loss is essentially twice the amount of SPY but profits for all the wins is 10 times that of using SPY and even more if spreads are used.
c. [21 day close no matter the situation] occurs and the put value is greater than what it was sold for. This is just nature of the beast and will happen at least at some point. These should be limited based on history and banking profits when they appear may be a way to stave off these events as much as possible, but they are a fact of how we invest and we need to be prepared for this situation to occur. Can’t always be winners
In summary, if using SPX Index options there is no threat of the options being exercised prior to the 21-day close date so the investor is always in control and the full amount of the underlying asset should not be looked at as capital risk. This creates a better situation and one that profits would be more readily served. Also, losses would be limited to the greatest extent.
3. Sorry to get ahead a bit. Will wait for Part 3 and see if I have questions then
4. Gotcha!
April 13, 2019 @ 10:52 am
You’re very welcome!
You’re spot on – in all 3 of those scenarios we’ll have lost money. While European options guarantee no early exercise, there is also no risk in having shares assigned – the risk was already realized by the underlying tanking, which can happen whether you’re at day 2, 13, or 44 in a 45 DTE trade. Being assigned is just a different way to have the losses realized. If the decision is made to hold a losing European-style trade beyond 21 DTE, you’ll be able to essentially defer the losses until expiration (provided there are no margin calls along the way as might occur in example “b”). The piper will have to be paid eventually.
Assuming VIX was around the historical average when the trade was placed in example “a”, this represents a drop of over 10% (~6.4% from the strike price and an estimated 4-5% OTM when the trade was opened for a 1SD put). A downturn of this magnitude would cause VIX to spike, making the following day’s trade measurably farther OTM for a 16 delta. The premium collected would also be greater, further adding to downside buffer. As you mentioned, this is why it’s important to always have a buffer [of cash / bonds / cashflow] to absorb initial losses while the subsequent returns begin to roll in. We don’t want future profits having to do double work of recovering portfolio losses while offsetting margin interest.
BigERN has a great visual of how this would play out in his post on making money options trading in Oct 2018. https://earlyretirementnow.com/2018/11/19/trading-like-an-escape-artist-made-money-in-october-trading-options-with-2x-leverage/. While the numbers would be different due to the duration tactics, the mechanics would be very similar – we’d take a loss on the initial underlying drop / VIX spike but then be compensated for it in future trades.
As for using SPX vs SPY, I’m not sure the math is adding up re: 2x loss but 10x upside. Assuming a SPY and SPX position were opened in parallel with identical greeks, in all scenarios the SPX will be 10x whatever the SPY is – naked, spread or otherwise. Trading SPX is essentially the same thing as trading 10 SPY contracts at a time.
In the example, if $1.5 is collected for the SPY put, $15 will be collected for the SPX put. If the underlying moves such that we’re sitting at a 5x loss, we’ll be out either $750 (7.5 * 100) or $7,500 (75 * 100). Keep in mind 0x loss = the initial premium received is lost – the trade breaks even and no money was lost from a portfolio perspective. A 1x loss = the initial premium received was not only lost but now we also “owe” that much – the portfolio realizes a loss equal to the amount of the initial premium received. 5x would continue this trend 5-fold to 750 and 7500 for SPY and SPX, respectively.
April 12, 2019 @ 7:50 pm
Sorry for the way this last reply is structured The comment area removed all of my tabbing and CRTN/Line feeds
May 8, 2019 @ 1:38 pm
This is a great series of articles, Spintwig! I have some questions and comments.
1. What’s difference between “per dollar of capital” (capital efficiency) and “per dollar of capital at risk” (risk efficiency)?
2. I had a problem with TT’s conclusions on the fourth table. They highlight the strangle in green, which suggests that is their winner. However, while SD is the lowest (31%), return (25%) is the median. If you divide return by standard deviation to measure a risk-adjusted return, then puts alone (100%) are the winner (0.837).
3. The fifth table looks decisive, but I wonder if it’s a bit misleading. Is it an apples-to-apples comparison? Average credit/day, which reflects potential profit, is 0.647 for 45 DTE and 0.171 for 15 DTE. Perhaps that 15 DTE strangle should be traded x4 to equalize? Also, the 45 DTE strangle is going to have double the risk due to overlapping positions. Maybe it needs to be halved in size? This table doesn’t show ROC or BPR, which could reveal differences. I’m reaching here, and I don’t think I’m doing it successfully. It’s amazing, though, how some explanations so favor the shorter DTE and others the longer. Like I said, to me this table seems pretty decisive in favor of longer DTE.
4. Close trades at 50% max profit or 21 DTE, whichever comes first and then place new trade immediately or wait until 45 DTE? Over the totality of TT’s research, it’s not clear to me when their backtesting results have reflected overlapping positions and/or when part of the backtesting interval holds the account in cash waiting for a particular trade day.
May 8, 2019 @ 10:56 pm
Thanks Mark, happy to hear you’re enjoying the series.
1. Capital efficiency is a measurement of return on capital (ROC) over time without regard for the risk being taken / volatility realized. Essentially a total P/L measurement charted in increments of 10 deltas. Risk efficiency is a measurement of returns per unit of volatility. Essentially a measure of risk-adjusted returns charted in increments of 10 deltas.
2. Agree. I interpreted it as: for a 39.0% reduction in returns (25 / 41) -1, we realize a 36.7% (31 / 49) -1 reduction in volatility, which is roughly a 1:1 trade off, though it agrees with puts being more efficient. Perhaps the confidence interval on the stats they used is coarse enough to make the difference statistically insignificant. I stopped trading calls so strangles are out of the equation for me.
3. That’s a great point about overlapping. It does’t say whether they opened one of each position every Monday, which would cause significant P/L differences as seen in the table, or one 45 DTE every 3 Mondays and one 15 DTE every Monday, which would be more apples to apples. If we put that study aside and use the next slide about cumulative P/L over trade duration, I think that has a cleaner argument for avoiding gamma risk by using longer dated options.
4. Based on the numbers I’ve experienced in my own trading, I suspect after they take profits they open another position immediately. Getting a ~3x ROC boost, as depicted by their annual P/L numbers, is consistent with my empirical findings of doing the same.The curious part is how they handle several positions hitting profit targets in a single day. If the mechanic was to simply open new positions after old ones were closed, the entire pool of available capital would end up being traded in a single day due to the influences of volatility.
Thoughts?
May 9, 2019 @ 9:19 am
Okay so risk efficiency is a subset of capital efficiency: just the latter categorized by delta of short strikes.
This, and most studies they do, would be greatly enhanced by inferential statistics.
WRT 3., are you talking about the “cumulative P/L of a 45-day 16-delta SPY strangle” graph? I don’t see how that pertains to a comparison between shorter- and longer-dated options.
WRT 4., do you mean several positions on different underlyings? The concentration effect does make sense even if they stay in cash, somehow, until IVR is higher.
May 9, 2019 @ 11:42 pm
Agreed
WRT 3. Yes. As the DTE approaches 0 the theta collected per day decreases significantly. This is the “why” behind the whipsaw chart in part 4. Shorter-dated options have a higher % of losers than longer dated options. There’s nothing magical about the option duration itself – it’s all about saying out of the 21 DTE range and collecting premium decay before 21 DTE is breached. It’s around the 21 DTE range speed (the gamma of gamma a.k.a. 3rd derivative of underlying spot price) increases and the option’s volatility crosses an inflection point of acceleration.
WRT 4. Not necessarily, but the principle remains. Different underlying would have different volatility profiles so this effect would be muted; it would just take longer to reach a state of all trades closing and opening on a single day. It’s essentially the same concept as “flapping” when architecting packet flow over a router or switch: a flow-control algorithm needs to be defined.
Tip: per your question in the other post, this is one of the reasons for introducing a delta variable (trading across deltas) when placing trades – it’s introducing diversity, yes, but more importantly it’s adding a little bit of variance into the equation to dampen this “flapping” effect. If a 15.5 delta position reaches a profit target but a 16.2 delta doesn’t, the problem has been mitigated without significantly changing the risk profile of the strategy.
May 10, 2019 @ 8:14 am
WRT 3. Are you using the shorter DTE on the 45 DTE trade P/L graph to simulate opening a trade with shorter DTE? The two are not the same. It would be interesting to look at a similar P/L graph for trades opened 15 DTE (or in the case of Part 4, intraday data could allow you to look at trades 4 DTE or less and I think the relevant comparison would be obtained by normalizing for [fractional] DIT / DTE).
WRT 4. I think the concentration effect is a problem because it eliminates temporal diversification.
May 11, 2019 @ 7:02 pm
WRT 3. I’m not using graphs, I’m inferring [due to lack of TT data] 🙂 I disagree that they are not the same; it’s a function of order entry.
Suppose I sell a 10 delta 260 SPY PUT 45 DTE and close it after 38 days. There is a specific risk profile associated with the position and trade (i.e. gamma) that’s directly correlated to the days remaining. I could re-open the exact same position after locking in profits and the reopened position could go on to become a loser.
Shorter DTE horizons don’t afford as much opportunity to turn what would be a losing trade into a winner by managing early.
Per the graph, notice the 5th-to-last bar which roughly represents 7 DTE (including non-trading weekend days). The slope of the final P/L is shallower than the slope of days 1-21.
WRT 4. Agree.
May 12, 2019 @ 11:31 am
When I say they’re not the same, here’s what I mean. The last five bars of the graph above would not look the same as the last five bars of a 7 DTE trade (which would basically be the whole trade).
If you want to compare PnL profiles, then what I think would be similar would be percentages. I would expect the first half of the above graph to the similar to the first half of a 7 DTE trade (might have to break down the total time into hours to get an equal number of bars). I think the last six bars of the above graph (6 / 33 = 2 / 11) would be similar to the last six bars of a 7 DTE trade if each bar were equal to: 5 business days * 6.5 hours/day = 32.5 hours / 33 bars ~ 1 hour per bar.
I would hypothesize that shorter DTE horizons offer as much opportunity to turn losing trades into winners by managing early–we just have to decrease the monitoring interval. For a 45 DTE trade, perhaps you look once daily; for a 7 DTE trade, perhaps we need to look roughly every 7/45 of a day, which is roughly every hour.
July 1, 2019 @ 2:36 pm
It’s worth noting that RobinHood front-runs their customers. And it’s a pretty good bet that you lose far more through that than whatever fees you’d be paying at, e.g., TastyWorks – known prices in a competitive market get optimized, whereas with hidden ones, there’s no motivation to do so. TANSTAAFL as usual – and doubly so for almost anything in the world of finance.
https://seekingalpha.com/article/4205379-robinhood-making-millions-selling-millennial-customers-high-frequency-traders
July 1, 2019 @ 5:21 pm
Thanks for visiting Ben —
This is true! As the saying goes: “if the product is free, you are the product.” Gmail and the rest of GSuite is an example, as is Facebook.
AlphaArchitect has an article about payment for order flow, too: https://alphaarchitect.com/2015/03/16/shedding-light-on-payment-for-order-flow-insights-from-emmet-peppers/
One of the posts in my drafts folder is on the business model of retail brokerages. This topic fits nicely.
July 28, 2019 @ 4:38 pm
Spintwig, have been poking around your site now for over an hour. Lots of good info. On issue of broker. On a large broker not mentioned here I was able to get a flat .50, even on 1 lot. Will not mention name as took months to get this deal and do not want them bugged by others trying to get same deal. I also have an IB acct and sometimes its .55 sometimes negative, and sometimes 1.38. I never am sure what price will get hence worked on another broker for predictable comms.
On Robinhood was not aware of the free. So the issue of trade price is big. On 1×1 trades the comms are tiny, $1 but on 8×7 Cals or Diags its $7.50. Their executions are the best have ever seen, sometimes on mkt orders to sell get the Ask. And getting the mid is maybe 35% of the time. I compliment them on this all the time. What about Robinhood? what percent of time do you get mid or better? On SPYs if its 1 penny spread then getting the bid on sale is fine. What is your experience on price execution there?
July 29, 2019 @ 3:22 pm
A few weeks ago I read about a broker offering a flat $0.50 per contract if the client asks for a negotiated rate. This post was well before I learned about this.
As for RH, I don’t have any firm data on price execution. I used them for maybe a month in 2018 before going to IB and unfortunately don’t recall the execution quality. The reason for the jump to IB was a promotional link that offered a lifetime discount on the first tier for margin interest – the first 100k is assessed the same rate as 100k-1M. This was fitting since access to cheap capital is part of my big-picture cash management strategy.
June 26, 2020 @ 4:06 am
>It’s worth noting that RobinHood front-runs their customers. And it’s a pretty good bet that you lose far more through that than whatever fees you’d be paying at, e.g., TastyWorks
Ben, RH isn’t the only broker that does this. tastyworks sells order flow to the exact same HFT firms as RH; check their disclosure. In fact, even “traditional” brokerages such as Fidelity and Schwab do the same.
October 4, 2019 @ 1:33 am
Hi Spintwig, great series!
Could I ask whether a similar strategy (naked put and/or strangle) could be implemented with far less capital <$5,000 by using SPY mini options (I think equivalent to only 10 shares) and no leverage?
Due to being a non resident I would have access to margin at only <8-9% interest rate, eating into my returns.
Cheers!
October 4, 2019 @ 6:00 pm
Hi Jack! Thanks for visiting.
It can indeed. If total notional exposure is of concern I’d start with spreads since it can potentially have a smaller capital/margin requirement. Also, I’d recommend Robinhood as the platform to implement your strategy since it’s fee-free. The dollar or two round-trip fees on other platforms will add up quickly when working with small numbers.
Worth noting, XSP has quite limited liquidity relative to SPY. I’d stick with SPY as the underlying due to the substantial liquidity it offers over comparable products. The tight spread on SPY will outweigh any tax inefficiencies (assuming you’re US based) from SPY not having Section 1256 treatment.
Hope this helps!
October 4, 2019 @ 9:04 pm
Hi Spintwig, sorry for spamming.
I think I understand what you meant by using a spread now and please correct me if I’m wrong.
By using a bull put spread I could kind of replicate the naked put strategy but I would limit the risk (up to the spread amount) and limit the payoff (short put premium minus long OTM put price). Is that right?
Is there an optimal setup for this spread? Still -16 delta for the short put and how far OTM should the long put roughly be?
Thank you so much!
Maybe you could write a segment on best strategies for beginners tight on cash/no margin 🙂
October 4, 2019 @ 9:13 pm
Hi Spintwig, sorry for spamming.
I think I understand what you meant now and please correct me if I’m wrong.
To replicate the naked put strategy, I could use a bull put spread thereby limiting my risk and capital requirements (to the width of the spread) and limiting my upside by the amount of premium minus long OTM put price.
If I may ask, do you know whether there is an optimal setup for this? Short put at -16 delta? How wide should the optimal spread be?
Thank you so much!
Maybe you could consider doing a post on optimal/replicable strategies for people tight on capital/ no margin 🙂
October 6, 2019 @ 12:23 pm
Exactly!
Great question. For SPY underlying, the answer is over at https://spintwig.com/spy-vertical-put-spread-strategy-performance/. All “standard” scenarios have been backtested and key performance indicators have been published. The 10D/5D spread held till expiration has the highest Sharpe ratio. Adjusting to 10D/2.5D or 16D/5D can increase profits at the risk of greater drawdowns and higher volatility.
That’s a great idea. Let me look into that.
October 4, 2019 @ 8:12 pm
Hi Spintwig, thanks for the reply! Yes unfortunately I am not US based so I don’t have access to Robin Hood, the cheapest brokerage I could get is probably tastyworks. Also, no preferential tax treatment.
May I ask why spreads would be more suitable? From your Part 1, I was under the impression that naked puts and strangles had a more suitable risk/reward profile.
If I may ask one more thing. What are your thoughts on doing poor man covered calls (long deep ITM leap possibly on an index and selling short delta 16 or 30 calls) could that be more suitable to my situation?
October 6, 2019 @ 12:33 pm
Ah. Interactive Brokers is where I do all my trading, though I’m not familiar with the rates for non-US folks. Sometime this month they’re releasing IBKR Lite (again not sure of target audience) which is designed for smaller accounts. tastyworks I hear is also a good choice; I’ve not used them myself so I can’t speak from personal experience.
Poor man covered calls on SPY generates losses overall. The SPY short call study at https://spintwig.com/spy-short-call-strategy-performance/ explores this in detail. If you’re using long LEAP as the “collateral” as opposed to long SPY, the losses will be even greater due to the decay of the LEAP.
May 22, 2020 @ 11:40 am
Hi, I was wondering if you still don’t manage losers, and if you don’t, do you immediately sell back your stock if you get assigned on a naked put and take whatever loss comes with it?
Many thanks,
O
May 23, 2020 @ 3:38 pm
Hi O.
It depends on the underlying.
For example, options on SPY tend to have better absolute and risk-adjusted retunes when managing early. That is, I wouldn’t hold any till expiration. I’d buy back losers for a loss at, say, 21 DTE. This has the added benefit of avoiding most assignment scenarios.
Meanwhile, options on T tend to have better absolute and risk-adjusted returns when holding till expiration. For losers on this underlying is it much more likely to experience assignment. When this happens, yes, I’d receive the stock then sell it back for a loss.
Some traders will hold the assigned shares then sell calls agains them but I personally don’t do that.
May 23, 2020 @ 3:33 am
Early execution:
In the setup explanation you probably have overlooked that it is assumed that early execution never happens.
Loss management:
I presume that it would hardly make a difference if you manage losses. That is because if an option moves far into the money, the win/ Loss chance moves towards 50% while delta moves to constant 1 . In reality however it is better for the nerves and it saves margin if you manage lossses.
May 23, 2020 @ 3:56 pm
Yes, early assignment is assumed to never occur. This writeup was due for a refresh, which I finished today. The studies referenced highlight this assumption in the methodology section. Specifically: https://spintwig.com/spy-short-put-45-dte-leveraged-options-backtest/#Moneyness
“Managing Positions” section has been refreshed to highlight the P/L benefits of using both a profit target as well as taking positions off early regardless of individual position P/L. The idea is to keep portfolio gamma lower which has broader implications for P/L and volatility.
August 13, 2022 @ 10:05 pm
Hi there,
Great article and great review at https://earlyretirementnow.com/2020/06/17/passive-income-through-option-writing-part-5/.
Karsten’s strategy is 1-3 DTE and survived the 2020 bear market which is super impressive.
Being an investor, our #1 priority is to preserve capital. Just wondering, would the 45 DTE also survive that environment?
Would love to get your thoughts and thanks in advance.
August 14, 2022 @ 4:40 am
Yes. If properly sized, the 45 DTE strategy would have also survived. Additional detail at: https://spintwig.com/how-to-trade-options-efficiently-part-3/#comment-8237