About this template
What are Defined Outcome ETF's?
Defined Outcome (also known as Buffered) ETFs are designed to target specific returns during outcome period. These ETFs typically will absorb the initial downside movement of the ticker, while also capping the upside. Outcome periods are typically quarterly, semiannually, or annually. A buffer ETF that has a 10% buffer and a 5% upside cap, you won't see any losses unless the underlying decreases by 10%. At that point, you should see a point for point loss. The maximum gain would be capped at a 5% gain. These buffers and caps are reset at the end of each outcome period. The annual cap on a quarterly period ETF is 4 x the quarterly cap, while the annual cap on a semiannual period would be 2 x the semiannual cap. I should point out that these outcomes will exist only if you own the ETF at the beginning of the outcome period, and own it at the end of the outcome period.
What's the difference between a Collar and these ETFs?
The upside part of the collar behaves similar to these ETF's. However the downside behaves the opposite. The downside of a buffered ETF behaves similar to a short put spread. You won't encounter any losses until the underlying breaches the short put, then your losses are capped where the long put is placed. You absorb the initial losses on a collar until the long put is reached.
What does a typical risk profile look like?
The Dashed line represents the price of the underlying, while the solid line shows the diagram of the ETF. This diagram shows the buffer would absorb the initial downside, the gain is capped.
What types of securities do these ETF's hold?
These ETFs hold FLEX options to create their positions. The positions consist of a long call spread with the short massively extremely deep in the money (SPY strike of only $2 or $3), with the short strike placed at the cap amount. The long call spread is paired with a long put spread where the long at the money, and the short placed at the buffer amount. The following is the list of securities that BALT currently holds. BALT is a quarterly buffered ETF with a 20% buffer. The cap documented in the prospectus for this ETF is 2.56% (before fees/expenses).
Here are the current holdings for BALT. The strike prices aren't the same as the listed options on SPY. The first thing I noticed was the strikes aren't the same as listed options for SPY. The long call, at the strike of 1.19, is very DITM. SPY closed on Dec 29, 2023 at 475.31 (the same strike as the long put). The 380.25 short put is 20% below the Dec 29th closing price of SPY. The long call is placed so that it provides close to 100 delta, and the short call is placed at the 2.56% cap of 487.48.
Can we make a bot that (generally) replicates one of these ETF's?
Sure, with some caveats.
- We can't go as deep ITM on the long call
- We can't be as precise with the strike prices
- We have to be cognizant of the potential early assignment on SPY
- We also have to guard against "no-bid" situations on the long options.
With that out of the way, let's convert the original structure to something we can trade. Let's go back to Dec 29th and use existing strike prices of listed options for March 28th expiration:
Long Call Spread
Long 1.19 C => Long 275 Call
Short 487.48 C => Short 487 Call
Long Put Spread
Short 380.25 P => Short 380 Put
Long 475.31 P => Long 475 Put
Embedded Inverted Straddle
The both of the legs of the long put spread are between the long call spread, which means the shorts of the spreads and the longs on the spreads are inverted.
Not a big fan of inverted straddles. Inverted straddles means that at least one of the legs is ITM. ITM spreads (especially deep ITM) increases risk of assignment and increases the opportunity for bad fills. We will address the inverted straddles later.
Long Call Spread -> Short Put Spread
There are two main problems with the long call spread. The first problem is the possibility of early assignment on an ITM long call, and the second problem is the bid-ask spread width of a deep ITM option. Knowledge of synthetic equivalencies tells us a long call spread is the same as a short put spread when the strike prices and expirations are the same. We will change the long 275 Call to a long 275 put, and the short 487 call to a 487 put. Note that now the 487 put is ITM. This leaves us with:
Long 275 Put
Short 487 Put
Short Put Spread -> Long Call Spread
Similar to the long call spread, the original short put spread has an ITM leg. We can also convert this from a long put spread to a short call spread. The long 475 Put becomes a long 475 Call, and the short 380 Put becomes a short 380 Call. Now the 380 Call is ITM.
Long 475 Call
Short 380 Call
This conversion solves the ITM 275 Call and the 487 ITM put. We now have an inverted straddle consisting of an ITM short 380 Call and an short ITM 487 put. We can simply revert the inverted straddle to a normal straddle by moving the long 380 call to a long 487 call the short 487 put to a long 380 Put.
We end up with:
Short Put Spread
Short 380 Put
Long 275 Put
Long Call Spread
Long 475 Call
Short 487 Call
Let's go back to this image.
There are 3 visible "bends" in the option diagram. The "Buffer" would be the 380 strike. The "Cap" would be the 487 strike. The bend between the "Buffer" and "Cap" would be the 475 (ATM) strike. The long put at 380 begins the buffer, while the long call at 475 is the point where the underlying starts to gain.
All of the strike juggling makes me think I should have been in a circus.
OK, so what do all of these changes mean.
- Potential of early assignment is mitigated by moving ITM strikes to OTM
- Bid-Ask spread of ITM options is reduced by rearranging ITM options to be OTM.
All of the strike juggling makes me think I should have been in a circus.
Now, on to the bot.
The dashboard
Bot complete o/a Feb 13 2023, with first positions opening Mar 1, 2023
Current Open Positions:
Closed Positions:
This bot is based on the ticker BALT ETF (Wealth Shield). Other buffer ETFs exist that are semiannual or annual outcome periods, different buffers and caps, and different symbols. Even though I had to change things for bot trading, I wanted to keep it as similar as I could to the original ETF.
This is a very capital intensive bot. This can be modified by moving the long put of the short put spread closer to the money. Moving the long put up will increase the cost of the long put, but will reduce the buying power. Moving the long put will have the added benefit of minimizing the risk of a "0-bid" situation.
The bot was originally designed to change positions on the 1st day of the 3d month of every quarter (Mar, Jun, Sep, Dec) in order to exit the position prior to the ex-dividend date that happens on the 3rd Friday of those months. I changed this to close existing positions and open new positions on the 2d Friday of those months. I did this to allow more of the short put/short call premium do decay.
If I were running this bot live, I would change the positioning of the short put. Using a "30% below the current price" would place the long put at the 340 strike, which would cut the amount of capital for the short put spread by 60% from ~ 10,500 to ~4,000. I also would consider increasing the upside cap to 4 to 5%, which would target a return of 16 to 20% per year.
I also probably would increase the buffer to around 15%. Only 3% of the 3 month periods since 1949 have experiences a decline of 20% or more. I would be willing to "self-insure" for that additional 5% loss.
