Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

SOXL - Leveraged Up The Wazoo

|Includes: Advanced Micro Devices, Inc. (AMD), SOX

Summary

Direxion Daily Semicondct Bull 3X ETF (SOXL).

A leveraged sector ETF.

With a sting in the tail.

Direxion Daily Semicondct Bull 3X ETF (NYSEARCA:SOXL)

In this post we'll take a look at SOXL, a leveraged ETF promising 3x the daily returns of the Philadelphia Semiconductor Index (^SOX).

What Investors Think They're Getting

A naive investor is likely to buy SOXL thinking that their position is equivalent to a 3x larger position in SOXX (an unleveraged SOX ETF). In option terms, they think they're getting three lots of SOX delta (though we'd usually refer to beta in an index context).

Leverage and Path Dependency

Because of the leverage in the product, SOXL returns are highly path dependent. I.e. the investor's P&L is impacted less by the level of the underlying index than by how it gets there.

An Example

To illustrate, let's consider an exaggerated example. We're cranking up the size of daily moves to amplify and highlight the difference between leveraged and large investments.

Day 1

On day one, investor A, an impoverished Millennial buys $100 of SOXL, hoping to get exposure to $300 of the SOX index. Investor B, his rich uncle buys $300 worth of SOXX, an unleveraged SOX ETF.

The SOXL ETF manager leverages up Investor A's $100 of capital to $300 of market exposure by buying a $300 SOX basket. (In reality, derivatives are used, but someone has to buy the exposure, so the end result is the same.)

The SOXX manager also buys a $300 SOX basket with investor B's capital. (Technically the stock is bought by an 'Authorized Participant' intermediary.) 

Let's assume that A and B's timing sucks and the SOX index falls -20% on the day. A loses $300*20% = $60 of capital. Investor B also loses $60.

The manager of the SOXL ETF has to sell $300 - $60 - $120 = $120 of stock at end of day to match B's day 2 capital ($40).

The manager of the SOXX ETF does nothing.

Day 2

On day two,  A's $40 of capital ($100 - $60) is leveraged up to $120 of SOX exposure.
Assume the Index bounces back +25% to its original level.
A's capital increases by $120*25% = +$40 to $80. B's capital grows from $240 to $300 - back where he started.

The manager of the SOXL ETF has to top up the fund's exposure to 3x A's $80 of capital, so buys another $80*3 - 125%*$120 = $80 of stock at end of day to match A's day 3 capital.

The manager of the SOXX ETF does nothing.

The End of the Path

So, after 2 days, with the index unchanged, Investor A has lost $20 and the SOXL fund has bought $300, sold $120, bought $80 of stock in a hurry in the last hour or two of trading. Investor B's capital is unchanged at $300 and the SOXX manager has done nothing beyond his initial purchase of $300 of SOX stock. 

What Investors Actually Get

The difference between returns for A and B is determined by the depth of drawdowns in the index between the initial investment and our end date. As you'll see from a zoomed out copy of my profile picture -

Random Walk
the depth of the drawdowns is a function of the volatility of the index. The higher the volatility, the more extreme the zigs and zags. So, in option terms, investor A is short volatility or the (made up) Greek vega.

If you consider the SOXL manager's re-hedges - selling into the close on down days and buying into the close on up days - Investor A is also short gamma.

Investor B is just long plain old boring delta (or beta in an index context).

You can see an illustration of the impact of volatility levels on returns in a table on page 3 of the SOXL prospectus (SOXL Prospectus). From the table, we can see that if the index were to rise +60% in twelve months with 60% realized volatility, Investor A's expected return would be $0, while Investor B would of course make $180.

Impact on AMD

We've seen from a previous post ("Who Trades AMD?") that AMD is over-represented in the SOX index, and that SOX index basket trading will have 6x the impact on AMD as on INTC, and 4x the impact on NVDA.

The SOXL ETF is about $700m in size. At a ~2% SOX index weight, the ETF manager will have $700m x 2% x 3 = $42m of AMD exposure. If ^SOX falls -5%, the ETF's capital will shrink to $595m ($700m x 85%), and the manager will have to reduce his AMD exposure to $35.7m ($595m x 2% x 3), selling -$6.3m of stock (or equivalent option trades).

An incremental sale of $6m of stock at the end of the trading day is more than enough to cause the stock to nose-dive - in a manner we've seen so many times in late trading.

Mitigation by Inflows

In periods when there's investor demand for SOXL, 'authorized participants' will approach the SOXL manager with new capital to be invested the the ETF. The purchase of leveraged exposure for this new capital will offset the re-hedge selling on down days and amplify buying on up days.

The manager can smooth out flows by letting the market bid up SOXL above NAV or sell down to a discount for a day or two, before meeting new demand by making (or breaking) ETF units.

Disclosure: I am/we are long AMD.