I took an interest in mechanics of leveraged ETFs when I knew about two new products by direxion with exposition in RSX. I faced a huge number of critics on such investment products, also lot of critics on these critics. Even I found some strategies based on myths and misunderstanding these and similar financial instruments. I cannot say that leveraged products are better than investment without leverage exactly. As I can't conclude that leveraged products are better or worse leveraged exposition on underlying benchmark. It depend from investment policy of everyone, everyone's preferences which are different, as it well known.
Anyway usually conclusions are based on insufficient data. This can explain why some myths are so popular and why there are to many misunderstandings. I point on effect of leverage. When we judge some investment product we compare it with related benchmark. Usually we either watch changes in final worth or study compound rate of return. At the practice (and at theory too) leveraged scheme can not beat its own unleveraged underlying very often. But it does not mean that future value of worth (expected value) suffer from leverage definitely.
To prove this let's perform simplest experiment based on Monte-Carlo Simulation. Why MC? Because it is
Assume, RSXs dynamics is based on Geometric Brownian Motion with growth 12% annual and volatility 30%. These numbers are much close to reality (at least close to some popular estimations). Let's study three investment strategies. First one: buy and hold RSX directly. Second: borrow money and buy and hold RSX with leverage 3x ether till end of some defined period or till "margin call" happens. Third scheme are purchasing of some leveraged ETF (or other product) which adjusts leverage every day to 3x.
Additional conditions: At second strategy when investor's worth falls below "zero" worth becomes equal "zero" (here we model limited liability kinda like shareholder's liability). At the third scheme let's define maximum daily change of worth as 100% from previous value.
Suppose we start a 500 days period (approximately, two years) with a portfolio consisting of equal amounts of capital in these three schemes (let it be 1 dollar in each strategy). What do we get after 2 years? Final worth in 1st, 2nd and 3rd strategies are 1.27 (+27%) 2.05 (+105%) and 1.84 (+84%). Expected worth of leveraged ETF beats and direct investment (1st strategy) and "buy and hold" scheme with borrowing (2nd).
But what about chances to beat market? So for each 7 from 10 time Leveraged ETFs were worse then benchmark purchased following 2nd strategy (borrow money)!!! There is 2 chances from 3 that unleveraged benchmark beat leveraged ETF and only 37 from 100 that benchmark will be better then 2nd scheme.
Results of this simple experiment (study) show following:
Anyway usually conclusions are based on insufficient data. This can explain why some myths are so popular and why there are to many misunderstandings. I point on effect of leverage. When we judge some investment product we compare it with related benchmark. Usually we either watch changes in final worth or study compound rate of return. At the practice (and at theory too) leveraged scheme can not beat its own unleveraged underlying very often. But it does not mean that future value of worth (expected value) suffer from leverage definitely.
To prove this let's perform simplest experiment based on Monte-Carlo Simulation. Why MC? Because it is
- easy (every one is able repeat this)
- illustrative (you can "touch" it and feel its mechanics)
- representative (much more data, actually here is artificial data)
Assume, RSXs dynamics is based on Geometric Brownian Motion with growth 12% annual and volatility 30%. These numbers are much close to reality (at least close to some popular estimations). Let's study three investment strategies. First one: buy and hold RSX directly. Second: borrow money and buy and hold RSX with leverage 3x ether till end of some defined period or till "margin call" happens. Third scheme are purchasing of some leveraged ETF (or other product) which adjusts leverage every day to 3x.
Additional conditions: At second strategy when investor's worth falls below "zero" worth becomes equal "zero" (here we model limited liability kinda like shareholder's liability). At the third scheme let's define maximum daily change of worth as 100% from previous value.
Suppose we start a 500 days period (approximately, two years) with a portfolio consisting of equal amounts of capital in these three schemes (let it be 1 dollar in each strategy). What do we get after 2 years? Final worth in 1st, 2nd and 3rd strategies are 1.27 (+27%) 2.05 (+105%) and 1.84 (+84%). Expected worth of leveraged ETF beats and direct investment (1st strategy) and "buy and hold" scheme with borrowing (2nd).
But what about chances to beat market? So for each 7 from 10 time Leveraged ETFs were worse then benchmark purchased following 2nd strategy (borrow money)!!! There is 2 chances from 3 that unleveraged benchmark beat leveraged ETF and only 37 from 100 that benchmark will be better then 2nd scheme.
Results of this simple experiment (study) show following:
- expected value of final worth does not suffer from leverage
- leverage cut rapidly your chances for success
No comments:
Post a Comment