Sanity check for a long-term leveraged investment strategy
Comments
-
The 'uncharted territory' you mention.
Looking at broad global indexes, the FTSE All-World had a peak to trough drawdown of 58% between autumn 2007 and March 2009, though after March 2009 when world governments got together and slashed interest rates to the floor and embarked upon massive money-printing, there was a rapid rebound, so it didn't stay below 50% for as long as 9 months, but the floor was about 18 months after the peak. You could not necessarily say that a rebound from the next crash would be as rapid as the last couple that we saw (from 03 and from 09 floors) because there is not necessarily the same capacity for global quantitative easing and interest rate reductions.
In 1989, Japan's TOPIX index for the Tokyo stock exchange peaked about 2885. In 1998, its low was 980. A decade and a half later, in 2013 it had a low of 872, but a high of 1302. In 2018, its low was 1415 (still 50% below its peak 29 years earlier) although its high was over 1900; it is 1735 today.
Similarly Japan's Nikkei 225 fell from 37,189 in Jan 1990, losing 35% over the course of that year and the subsequent decade from 1991 in Japan was known as the 'lost decade' due to the effects of the gradual collapse of the asset price bubble. After the earthquake in 2011, the Nikkei was at 8160.
So, single country indexes are best avoided. This is not to suggest the US has a price bubble due to collapse to the same extent that Japan did, and the constituents of its index are more varied than Japan's at the time. However, Japan was once the biggest world stockmarket and now it isn't. Faith in the S&P500 may to some extent be similarly misplaced; a broader index would probably be better, even if S&P ETFs are cheap to trade.0 -
Hi bowlhead99,
I've done my research on ETFs suitable for this strategy. I applied the below filtering criteria on all equity ETFs, and only 6 ETFs ended up in my list.
Expense ratio <= 0.8% (filtering out expensive funds)
Number of holdings >= 500 (physical replication broad market index fund)
Option volume >= 1,000 (filtering out funds with very bad option liquidity)
The 6 ETFs are:
SPY, SPDR S&P 500 ETF TRUST
EEM, ISHARES MSCI EMERGING MARKET
IWM, ISHARES RUSSELL 2000 ETF
EFA, ISHARES MSCI EAFE ETF
VWO, VANGUARD FTSE EMERGING MARKE
IVV, ISHARES CORE S&P 500 ETF
Excluding the emerging market and small-cap funds, there's only 3 left. Two of them are S&P 500 trackers, and the the other one is a MSCI EAFE tracker.
None of them on their own can represent the global market. My best idea would be combining two ETFs, SPY and EFA. Holding 60% of SPY and 40% of EFA, the portfolio is fairly close to the MSCI World index.
The only problem is it needs a lot more cash to start with. Because SPY is trading at $325.6, this portfolio would require nearly $28,000, and rebalancing can be very hard.0 -
Thank you. I posted the previous one before I saw this new post. Your comment on this issue is a very valuable input for me.bowlhead99 wrote: »The 'uncharted territory' you mention.
Looking at broad global indexes, the FTSE All-World had a peak to trough drawdown of 58% between autumn 2007 and March 2009, though after March 2009 when world governments got together and slashed interest rates to the floor and embarked upon massive money-printing, there was a rapid rebound, so it didn't stay below 50% for as long as 9 months, but the floor was about 18 months after the peak. You could not necessarily say that a rebound from the next crash would be as rapid as the last couple that we saw (from 03 and from 09 floors) because there is not necessarily the same capacity for global quantitative easing and interest rate reductions.
In 1989, Japan's TOPIX index for the Tokyo stock exchange peaked about 2885. In 1998, its low was 980. A decade and a half later, in 2013 it had a low of 872, but a high of 1302. In 2018, its low was 1415 (still 50% below its peak 29 years earlier) although its high was over 1900; it is 1735 today.
Similarly Japan's Nikkei 225 fell from 37,189 in Jan 1990, losing 35% over the course of that year and the subsequent decade from 1991 in Japan was known as the 'lost decade' due to the effects of the gradual collapse of the asset price bubble. After the earthquake in 2011, the Nikkei was at 8160.
So, single country indexes are best avoided. This is not to suggest the US has a price bubble due to collapse to the same extent that Japan did, and the constituents of its index are more varied than Japan's at the time. However, Japan was once the biggest world stockmarket and now it isn't. Faith in the S&P500 may to some extent be similarly misplaced; a broader index would probably be better, even if S&P ETFs are cheap to trade.0 -
This needs very little thought at all; gamblers or scoundrels invest with borrowed money. You can convince yourself that everything will be alright, but the costs and risks of failure are seldom correctly evaluated. Greed reinforced by numerology can be very dangerous in personal finance and it should be eliminated from any strategy. Leave option trading up to the professionals.“So we beat on, boats against the current, borne back ceaselessly into the past.”0
-
People who invest before pay off a mortgage is also investing borrowed money, but what's wrong with that?bostonerimus wrote: »This needs very little thought at all; gamblers or scoundrels invest with borrowed money. You can convince yourself that everything will be alright, but the costs and risks of failure are seldom correctly evaluated. Greed reinforced by numerology can be very dangerous in personal finance and it should be eliminated from any strategy. Leave option trading up to the professionals.
Leverage is double-edged sword. Used properly, it will do more good than harm.0 -
The risk associated with a mortgage is far less than options trading, but I do take your point which is why I made extra mortgage payments in lieu of some of my fixed income allocation in my portfolio. I’m now mortgage free, which is boring but also highly liberating.“So we beat on, boats against the current, borne back ceaselessly into the past.”0
-
I have a few more comments ... But really, you need to be modelling how your trade plan would have performed with historic data (as well as attempting to make some allowances for how the future won't exactly resemble the past). I'm just offering a few generalizations — and if you haven't thought of some of them, then consider that there are probably plenty more I've missed.
Combining new money and the proceeds from selling options that are to be rolled over makes sense from an operational point of view. (For one thing: it ensures you buy as many whole option contracts as you can with your total available capital.) But it's very misleading in terms of the theory of how your strategy is performing.
Why do I say that? Because I get the impression that you think that, so long as you are able to maintain your exposure to the underlying throughout a crash until the recovery, you haven't lost anything. And this is completely untrue. The comparison should always be with just buying the underlying (S&P 500), and if you did that, during a crash all your new money would be used to buy more of the underlying at "for sale" prices. With your options strategy, you instead have to use some or all of your new money to maintain your current exposure; so there is an opportunity cost. You are effectively trying to cover up the fact that your existing position in options is losing money by mixing it with the new money.
The more clear-sighted way to look at it is to separate (logically, not operationally) the handling of new money and roll overs. When you look at it like this, you have 2 choices when it's time to roll over an option that's losing money: keep to your target leverage (by reducing exposure) or maintain exposure (by increasing leverage). (Except that, in extreme cases, the latter may not be possible.) That's the choice to consider.
The way your plan mixes old and new money has the effect that how aggressive you are depends on the relative sizes of the option positions already held and of the new money being added. When the new money is relatively large, compared to what's already invested, it is less aggressive, because the exposure the new money can buy, while sticking to the target leverage, would in most market conditions be enough to cancel out the reduction in exposure from rolling over losing contracts into reduced exposure to maintain target leverage. But as the invested capital becomes larger relative to new money, the new money can't paper over losses so easily, and you could easily start ramping up the leverage. So this strategy is more likely to blow up badly after it's been going for some years. (Did you know that? If not, that's why you should have been modelling it.)
Is there any sense a strategy becoming more aggressive as you have more invested (relative to the size of new cash)? I don't think so. This is another reason to separate (logically) the handing of roll overs and new cash. I.e. to have a rule about how to handle roll overs which doesn't refer to what new money is available.
The more aggressive approach, of always maintaining exposure (if at all possible), is IMHO pretty mad if you attempt to follow it with no limit. Ultimately, you could end up buying far out-of-the-money options, waiting until they have nearly expired, still far out-of-the-money, and having lost most of their value when you bought them, and then reinvest the dwindling proceeds into new, longer-term, out-of-the-money options. Does that make any sense? If not, where do you stop?
In your plan as stated, there is no stopping digging when you're in a hole. (There is only a variable level of aggressiveness, increasing as you have more capital invested.)
I think any experienced trader (which I'm not — I just know enough to know I shouldn't be trading) will say there is a point where you have to back off and reduce exposure. You can't sanely always keep doubling down. (It's an appropriate thread title.)
Of course, we can all see why you don't want to maintain leverage by reducing exposure. Because then you don't get the 2X returns of the underlying (less the costs of using options) that you were aiming for. But the lesson is that leverage doesn't do exactly what you might want it to do. All it does is give you returns with a different shape from the underlying, which may or may not be higher. And, however you do it, with more ways it just might blow up in your face.
When considering how the effects of increasing leverage after a market fall, you'll note that (for instance) an at-the-money option is more expensive than a deep in-the-money option (for the same level of exposure), because the former is a 1-way bet, the latter a 2-way bet. So you are spending more on the time value of the option (which you then allow to decay as you hold the option, hoping that the return from the underlying compensates you for that loss). But in addition, after a market crash, volatility and therefore option prices are likely to rise; making increasing leverage even more expensive. All this makes doubling down more costly, and more risky. And this information should be fed into your modelling. (I think I already mentioned that you need to model this.)
But without modelling, we can say: using increased leverage during a crash adds risks, so you may be unable to stick it out until the recovery. But it also adds costs, so, even if you are able to stick it out until the recovery, the gains from staying fully invested may not be enough to cover those increased costs.
And a slightly different point: previously, I said most years would be good for this strategy. I was thinking that most years have no big crash. However, even aside from crashes and extreme events, in a mild year, with low but positive returns for an underlying index, the returns from an option strategy will be lower than the index, and could be a small loss, due to the cost of rolling (e.g. selling a 24-month option to buy a 12-month option). This alone, given many are projecting low returns from equities going forward, is a pretty good reason not to buy options now, when you could just buy the underlying index instead.0 -
under_western_skies wrote: »
But in addition, after a market crash, volatility and therefore option prices are likely to rise; making increasing leverage even more expensive. All this makes doubling down more costly, and more risky.
Tell that to Nick Leeson.“So we beat on, boats against the current, borne back ceaselessly into the past.”0 -
I could do the modelling, but I didn't do it, because I don't have much faith in financial modelling. Changing a few parameters can make a completely different outcome, and many of them (future inflation rate, interest rate, etc.) are predictions and expectations rather than facts. I'm pretty sure most fund managers have done their modelling, but did this prevent them from losing money? I don't think so.under_western_skies wrote: »But really, you need to be modelling how your trade plan would have performed with historic data (as well as attempting to make some allowances for how the future won't exactly resemble the past).
Instead of wasting time to fool myself with those seem prefect strategies from back tested models and then fail in the real world, I'd rather make my plan based on what is known, and then deal with the unknowns when they become known.
No, I don't expect "haven't lost anything". In fact my plan will decrease the leverage after the market has gone down, and increase the leverage while the market is going up. If you look at the flow chart, you will see that I'll always try to roll over to a new strike that's half of the underlying asset price. This means the leverage level get reset back to 2:1 each time the option is rolled over. Since this is a strategy to "buy high and sell low", it certainly will not maintain my exposure, and after the recovery there is a big chance I would have lost money. However, if the market is going up year after year, this strategy will also return more than twice of the underlying asset's return. One thing I know for sure, there's more years the market was up than it was down, and it's very likely to remain this way.under_western_skies wrote: »Combining new money and the proceeds from selling options that are to be rolled over makes sense from an operational point of view. (For one thing: it ensures you buy as many whole option contracts as you can with your total available capital.) But it's very misleading in terms of the theory of how your strategy is performing.
Why do I say that? Because I get the impression that you think that, so long as you are able to maintain your exposure to the underlying throughout a crash until the recovery, you haven't lost anything. And this is completely untrue.
Buy the "for sale" underlying asset after a crash will only work if it remains "for sale" for long enough while I'm waiting for my payday.under_western_skies wrote: »The comparison should always be with just buying the underlying (S&P 500), and if you did that, during a crash all your new money would be used to buy more of the underlying at "for sale" prices.
Actually, my plan is to reduce the leverage (roll to a lower strike) after a market downturn. This way I can survive a long lasting bear market. If my intention is to maintain the same exposure, I won't need much new money anyway, because the cost of roll over to the same strike in a later date isn't going to be high. Maintaining the same exposure is only a compromise if I must roll before the expiry and don't have enough spare cash to deleverage.under_western_skies wrote: »With your options strategy, you instead have to use some or all of your new money to maintain your current exposure; so there is an opportunity cost. You are effectively trying to cover up the fact that your existing position in options is losing money by mixing it with the new money.
My choice is clear, I will keep the target leverage. I will only maintain the exposure if I have no other choices.under_western_skies wrote: »The more clear-sighted way to look at it is to separate (logically, not operationally) the handling of new money and roll overs. When you look at it like this, you have 2 choices when it's time to roll over an option that's losing money: keep to your target leverage (by reducing exposure) or maintain exposure (by increasing leverage). (Except that, in extreme cases, the latter may not be possible.) That's the choice to consider.
In the event that I had no option but to keep the exposure, the possibility of market not going up at all in 2 years (LEAPS can last for 2.5 years) is slim. The chances are, as long as I don't put all eggs in one basket (one country, one sector, etc.), this isn't going to happen.
That's the next part of the plan, which isn't relevant yet. But since you've asked, I'll answer. By the time the portfolio is big enough, I will be able to sell some contracts to fund the roll if it's necessary.under_western_skies wrote: »The way your plan mixes old and new money has the effect that how aggressive you are depends on the relative sizes of the option positions already held and of the new money being added. When the new money is relatively large, compared to what's already invested, it is less aggressive, because the exposure the new money can buy, while sticking to the target leverage, would in most market conditions be enough to cancel out the reduction in exposure from rolling over losing contracts into reduced exposure to maintain target leverage. But as the invested capital becomes larger relative to new money, the new money can't paper over losses so easily, and you could easily start ramping up the leverage. So this strategy is more likely to blow up badly after it's been going for some years. (Did you know that? If not, that's why you should have been modelling it.)
Remember it? I intend to keep the leverage ratio at 2:1, therefore I'd have to reduce my exposure after a crash. Right now, my money can only buy me very few contracts, it's not economical to sell 3 and buy 2 contracts and end up with a lots of spare cash, therefore use my income to fund the roll is better. But when I have more invested, sell some contracts to fund the roll of others will become practical. Not only the proceeds from selling can be used to fund the cost of roll over, sell some contracts would also naturally reduce the leverage.
Of course, the sell will only happen when the market is down. Under good market conditions, the underlying asset price goes up, the leverage ratio would have gone down. The roll over will increase the leverage back to 2:1 and release some cash from the old contracts. The cash will be used to fund the cost of the roll. Cash from income will also be used to fund the roll if the market didn't go up enough. Ended up with spare cash? No problem, just buy more shares or option contracts.
I know too well that maintaining exposure will kill me. I wouldn't touch it.under_western_skies wrote: »Is there any sense a strategy becoming more aggressive as you have more invested (relative to the size of new cash)? I don't think so. This is another reason to separate (logically) the handing of roll overs and new cash. I.e. to have a rule about how to handle roll overs which doesn't refer to what new money is available.
The more aggressive approach, of always maintaining exposure (if at all possible), is IMHO pretty mad if you attempt to follow it with no limit. Ultimately, you could end up buying far out-of-the-money options, waiting until they have nearly expired, still far out-of-the-money, and having lost most of their value when you bought them, and then reinvest the dwindling proceeds into new, longer-term, out-of-the-money options. Does that make any sense? If not, where do you stop?
By keep the leverage ratio at 2:1, I don't see the need of a stopping point. A well diversified portfolio isn't going to fall 50% and then stays below 50% of the recent peak for years. Ironically, this strategy will actually benefit more from a slow recovery process spanning multiple years than a quick recover jump back to the pre-crash level.under_western_skies wrote: »In your plan as stated, there is no stopping digging when you're in a hole. (There is only a variable level of aggressiveness, increasing as you have more capital invested.)
I'm not an experienced trader. In fact I'm not even a trader. I hate trading, it benefits only the middleman. I make plans, and then execute them. Doubling down isn't a part of my plan.under_western_skies wrote: »I think any experienced trader (which I'm not — I just know enough to know I shouldn't be trading) will say there is a point where you have to back off and reduce exposure. You can't sanely always keep doubling down. (It's an appropriate thread title.)
This will not happen in an efficient market, because one can arbitrage by buying the cheaper ITM option and selling the more expensive ATM option, make a risk free profit equal to the difference between the premiums plus the different between the strikes minus the bid-ask spread and commissions.under_western_skies wrote: »When considering how the effects of increasing leverage after a market fall, you'll note that (for instance) an at-the-money option is more expensive than a deep in-the-money option (for the same level of exposure), because the former is a 1-way bet, the latter a 2-way bet.
The good thing about an efficient market is that those factors are priced in the option's premium, therefore now options are cheaper than they were. (current IV is at very low point comparing to the IV in the past 20 years).under_western__skies wrote: »And a slightly different point: previously, I said most years would be good for this strategy. I was thinking that most years have no big crash. However, even aside from crashes and extreme events, in a mild year, with low but positive returns for an underlying index, the returns from an option strategy will be lower than the index, and could be a small loss, due to the cost of rolling (e.g. selling a 24-month option to buy a 12-month option). This alone, given many are projecting low returns from equities going forward, is a pretty good reason not to buy options now, when you could just buy the underlying index instead.0 -
OK, so I was meant to know that the flowchart you drew was only about the initial stage, and that it would change at some unspecified point when you have more invested. It would make things clearer if you'd drawn a complete flowchart for your full plan.
The parts of the plan I was calling "aggressive" or "doubling down" are where you maintain exposure by rolling over into an option with a strike price higher than half the underlying or by postponing roll over. You now say that you "will only maintain the exposure if I have no other choices", but you always have a choice. The less aggressive form of your plan (which is also simpler, and doesn't implicitly change when you have more invested) would be:
1) At a fixed interval, take the new money available for investment.
2) Sell any options held with < 12 months to expiry (blindly, i.e. completely regardless of prices).
3) Use the money from (1) and (2) to buy as many options as you can, of the contract with an expiry just over 24 months and a strike price just under half the price of the underlying.
(That description omits the (uncontroversial) part about temporarily parking any money not sufficient to buy a whole option contract in the underlying.)
That is what I would call the less aggressive version of your plan. It is less liable to blow up in extreme conditions. (I still wouldn't recommend it.)
You don't want to model because you wouldn't believe the results blindly. Quite right about not believing them blindly; the point is not to do that, but to be able to talk about ways in which a strategy could do well or badly, and also to try out ideas about what would happen if market conditions changed in various ways.
Yes, but your strategy will underperform by a larger amount in the years the market falls than it outperforms in the years the market rises. So what are the chances of it outperforming overall? You simply cannot answer this without modelling (or some other formal method).One thing I know for sure, there's more years the market was up than it was down, and it's very likely to remain this way.
It is not answered by claiming that your expected return must have gone up because you are taking on more risk and the market is efficient. If that were the case, then a good betting method would be to always bet a little more than the Kelly criterion recommends. (In any case, the market is only efficient-ish.)
I didn't express that very well. I meant that the time-value part of the option price is lower for in-the-money than for at-the-money.This will not happen in an efficient market, because one can arbitrage by buying the cheaper ITM option and selling the more expensive ATM option, make a risk free profit equal to the difference between the premiums plus the different between the strikes minus the bid-ask spread and commissions.
E.g. if the underlying trades at 2000, so an ATM option has a strike price of 2000, and if an ITM option has a strike price of 1000; then the market price of the ATM option is the same as its time-value, and the market price of the ITM option is 1000 + its time value. The ITM option will have a higher price, but the difference will be less than 1000.
Well, it's a theory. But then future implied volatility is unknown. Current IV is low because markets have been relatively calm recently. If they become less calm, IV (and hence the cost of holding options) will rise. So if markets see-saw up and down a lot, holding options long-term will become more expensive. But will you be compensated for that? That depends onThe good thing about an efficient market is that those factors are priced in the option's premium, therefore now options are cheaper than they were. (current IV is at very low point comparing to the IV in the past 20 years).
(A) The overall compound returns from the underlying being high enough to compensate.
And the theories about expected returns being low now are not directly related to IV.
They are based on the idea that total returns = earnings yield + earnings growth + speculative premium (i.e. change in P/E ratio). (With the speculative premium either averaging 0%, or reflecting a reversion to an average P/E — depending on whose theory you're reading.)
Could there be some indirect relation between these ideas and IV? I've no idea. But unless there is, low IV shouldn't provide any comfort.
(B) The exact path that markets take to that overall return.
This is something that investors who just hold the underlying for the duration don't have to worry about. But when you use options, you do. Because your exposure to the market will be expanding or contracting as the market moves. You will be incurring higher or lower costs depending on the level of IV. And the exactly times you happen to have picked for roll overs can sometimes be crucial.
Clearly you are aware of most of the factors mentioned in the previous paragraph. These are the factors which IMHO make your plan in many ways more trading than investment. This is not to imply it's as bad as more hyperactive or discretionary trading strategies.0
This discussion has been closed.
Categories
- All Categories
- 343.2K Banking & Borrowing
- 250.1K Reduce Debt & Boost Income
- 449.7K Spending & Discounts
- 235.2K Work, Benefits & Business
- 608K Mortgages, Homes & Bills
- 173K Life & Family
- 247.9K Travel & Transport
- 1.5M Hobbies & Leisure
- 15.9K Discuss & Feedback
- 15.1K Coronavirus Support Boards