The momentum factor is one of the most frustrating forces at work on our portfolios. Frustrating because it can be a potent return enhancer – yet it has historically been difficult for an ordinary investor to actually exploit.
Momentum is one of the mighty return premiums that sit like gods in the pantheon of portfolio power-ups.
By tilting towards equities that are especially risky, or mispriced (because humans are serial dunces) these premiums have historically unlocked the door to market-beating returns.
Momentum has returned some of the biggest rewards of all: 9.62% annually in the US between 1927 and 2014.
What’s more, it’s been shown to exist in the UK since the Victorian era, and just about everywhere else bar Japan.
So what is momentum?
Momentum is the financial equivalent of a pitch invasion, where everybody crowds around the victors, hoists them shoulder high, invites them home to meet their mam, starts to dress like them, buys their fragrances, asks their opinion on the West Lothian question, and so on.
Success leads to greater success. Meanwhile the losers are shunned, despised, spat at, set upon by dogs…
Put more plainly, momentum is the tendency for winners to keep rising and losers to keep falling over periods of 12-months or less.
A momentum strategy buys the winners while selling or shorting the losers.
Why does the momentum premium exist?
Unlike with the value premium or the small cap premium, few researchers think that the momentum premium is a reward for taking greater risks.
Instead, momentum is about exploiting human weakness – specifically our inability to respond to news quickly and an inclination to under react or overreact to new information.
The idea is that lackadaisical investors digest news over the space of months, rather than incorporating it into the share price in nano-seconds, efficient market-style.
- Impulsive types might drive prices way beyond fair value while the sluggish might cling on to losers for a while before being forced to concede defeat further down the line.
- Momentum may also be partly due to forced selling from active funds that fall out of favour, or leveraged investors who face margin calls in bear markets.
- Yet another cause may be forced buyers who bid up prices further as they try to escape a short squeeze in a bull market.
Regardless of the explanation, the momentum premium has not been arbitraged away in the 20 years since its discovery – although it has weakened to a 6.3% annual return in the US since 1991.
Diversification
What makes momentum even more valuable is its low or negative correlation with other premiums.
Finding negative correlations is the Holy Grail of asset allocation. Holdings that act as counterweights can reduce your portfolio’s volatility when markets are turbulent.
Momentum is particularly interesting to value investors as the two factors have historically been negatively correlated (-0.27% in the US from 1927-2013).
Of course, like all the return premiums, momentum can fail to deliver.
It dished up an unsavoury annualised return of -1.3% a year from 2004-2013, which is a timely reminder that trying to capture return premiums is not a game for the fainthearted.
The momentum everyone’s been waiting for
The biggest problem with momentum is that it’s like a rare herb, with fabled medicinal qualities, buried deep in the Amazon jungle. It’s well known but very difficult to get hold of.
Currently there are only two momentum index trackers available in the UK, and they’re mere hatchlings with no track record to speak of.
Both are available in ETF form:
- iShares MSCI World Momentum Factor ETF – OCF 0.3%, ticker IWMO
- db X-trackers Equity Momentum Factor ETF – OCF 0.25%, ticker XDEM
Both ETFs cover the developed world including the UK and apply a formula to identify the highest momentum equities in the MSCI World index over the last 12 months. The equities that are judged to be on fire go into the ETF.
I’ll compare the chops of these two ETFs at a later date, but it’s important to understand that you shouldn’t just plump for the cheapest product and be done with it when it comes to risk premiums.
FTSE All-Share trackers tend to be much alike, whereas there are many different ways to slice a risk factor. Some products may be better than others at creaming off the extra return.
Make sure you understand:
- How the tracker intends to capture the risk premium.
- Compare those rules with independent definitions of the factor to try to assess whether it’s going to do a good job.
- How diversified is the product? Is it so concentrated that it’s more like a bet on certain industries?
- Is the index independent of the product provider?
- Is the tracker available at a reasonable cost?
- Does it have a good track record in comparison to other similarly styled products and indices? (Obviously that’s impossible to gauge for these new products).
If the product provider isn’t transparent about how its flashy smart beta tracker works then forget about it. Also, don’t rush into new products. They often need several months to bed in, so hold your fire until those bid-offer spreads settle down.
Multi-factor options
Given the proliferation of risk factors and firms racing to provide smart beta ETFs to track ’em, most of us will probably find it a lot easier to invest in a multi-factor ETF that combines value, momentum and other factors in one handy package.
If so, then the Amundi ETF Global Equity Multi Smart Allocation Scientific Beta UCITS ETF not only has a name longer than the average Welsh town but also attempts to capture momentum in its dragnet.
Sadly this tracker is only available on the NYSE Euronext Paris for now, but it’s probably only a matter of time before it, and other looky-likey products, are available on the LSE.
Dimensional Fund Advisor fans have no need to fret though as DFA funds already include a momentum screen as part of the service.
DIY momentum
You can capture some of the momentum premium by making sure you don’t rebalance too frequently.
Keep your moves to once a year, or even stretch it to every two years if you can. Or, if you use threshold rebalancing then keep your trigger points relatively slack (for example rebalance when an asset class moves 20-25% from target).
Take it steady,
The Accumulator
Comments on this entry are closed.
“The momentum factor”: this absurd noun “factor” is an invasive weed. I see it everywhere; it seems usually to mean nothing at all. Its verb form “factoring in” is even more loathsome.
P.S. Momentum isn’t a force; whether a “momentum factor” is a force, GOK.
Carry on!
@dearieme although I usually lose a usage and abusage of English debate with you I’d say TA’s use here is okay – momentum is one of the factors in his strategy which presumably has other strands that he hopes will give him a decent return, resulting in a middle finger signal to The Man earlier than otherwise.
Now as for using momentum as a way to get ahead, well, I have excellent experience of that in the dotcom boom. I marched all my men up to the top of the hill, whereupon it all went wrong very quickly.
Using momentum presumably involves the element of market timing in knowing when to get off the packed bus before everyone else has the same bright idea…
@TA
Thanks for the thoughtful post.
Quote
Regardless of the explanation, the momentum premium has not been arbitraged away in the 20 years since its discovery – although it has weakened to a 6.3% annual return in the US since 1991.
Unquote
Suppose momentum is unusual in the chartists’ armoury, in that the more people that seek and use momentum, the more it self reinforces.
Prices go up because they are going up and vice versa.
As ermine says above it is knowing when to get off the bus, as the first doubts among investors begin to creep in.
I read somewhere recently that if the name of a fund is not decipherable & you cannot immediately grasp its aim OR the manager appears to be trying too hard to impress/differentiate then swerve it.
“Amundi ETF Global Equity Multi Smart Allocation Scientific Beta UCITS”…….. so have to ask what is “Scientific Beta” about?
As an aside it is amusing that the use of the word momentum
= mass * velocity (speed or rate of change)
has come to be accepted so widespread in the investors’ dictionary.
Share Prices have no mass as far as I can judge.
It is a reassuring word for trend followers since to overcome momentum the word suggests we would have to apply an opposite largeish force to stop the momentum, without which share prices would continue on indefinitely rising or falling.
In practice share prices can turn on a sixpence, which would not happen if they really did have mass.
Four questions come to mind with ETFs:
1 Why are financial companies so keen to offer ETFs if the fees are genuinely as low as claimed?
2 Why do some tracker ETFs underperform their index by a substantial percentage? I currently just own 1 ETF. If I add underperformance to the fees, it’s costing me over 1% a year over the last 5 years. That just tracks a recognised index, not some momentum index.
3 This is less common, but in some cases, an ETF tracks the index & it’s difficult to work out where the dividends go to, considering the ETF is not tracking a total return index and not paying out dividends which match those expected.
4 If the ETF is using “swaps” and the counterparty is part of the bank which offers the ETF, guess who will lose out.
I shall continue to select good Investment Trusts in preference to ETFs.
I agree with Magneto.
‘Momentum’ actually conveys a quite inappropriate meaning.
I suggest ‘Bandwagon’ investment. That term seems to encompass the essence of it, including the risk.
@Jeff
1) Because they don’t have to worry about keeping track of who owns the units, or have to pay people to to run them.
2) Because some only sample the index, and some can get diddled slightly during the (rare for cap-weighted ETFs) occasions they rebalance. http://www.morningstar.co.uk/uk/news/66096/the-total-cost-of-etf-ownership.aspx
Pick Vanguard ETFs for best behaviour. 🙂
3) Possibly witholding fees. (e.g. Lyxor ETFs IIRC)
4) You, but again most ETFs have plenty of collateral. Besides, most are physical these days. (Securities Lending is relevant but they get collateral.)
I hold both ITs and ETFs.
A great comment thread. Subscribe me!
@magneto, perhaps in this context metaphorically speaking mass = investor exuberance & velocity = speed at which people pile in to invest in a stock/tracker/ETF. :()
@Jeff
These would be my answers:
1 To compensate for outflows out of their active funds + to make a huge sum of money with stock lending + SWAP activities.
2 Tracking difference can have multiple reasons. Usually the TD equals the TER (- stock lending gains in some cases). Some SWAP ETFs charge a SWAP fee on top.
3 Tax
4 The investor. Same with stock lending – you take the risk & the ETF company keeps most of the reward.
I shall therefore continue to run my own index of individual stocks.
aren’t all factors fickle? 🙂
all interesting theory, but the only bit i’m inclined to act on is TA’s last point, about being slow to rebalance in the hope of benefiting from momentum while you’re waiting (or at least, avoiding being disadvantaged by going in the opposite direction to momentum).
on ETFs, let’s not get too bogged down in the issuers’ motives – except perhaps in the case of synthetic ETFs, where IMHO a bit more suspicion is warranted. for most mainstream kinds of investment, you can now get tracker funds or (physical) ETFs which are a lot cheaper than the nearest IT – that’s the major advantage. also, it is a lot easier to understand how a tracker will behave than it is to know what approach an active manager will adopt and how successful it will be.
Hmm, you set me pondering with the comment at the end about rebalancing frequency. In cases where one is in accumulation phase with regular cash injection from income or divided reinvestment into the portfolio, is there any advantage to remaining un-invested? I can see the logic for having a generous threshold for top-slicing, but is the strategy to rebalance the poorer-performing components still desirable on a regular basis?
Ha ha. Thank you the pedant’s brigade 😉 And now for some investing…
Once the new ETFs have settled down a bit, I think I’m highly likely to invest in momentum due to its negative correlation with value, and low correlation with other factors / forces / styles / anomalies / premiums / phenomena / ephemera…
@ Jeff –
1. Because the popularity of ETFs means they can’t afford to ignore them. That said, steer clear of active ETFs, leveraged, inverse, sector specific, and most country specific. Here’s a piece on the specific risks you allude to: http://monevator.com/synthetic-etf-risks/
2. I second Greg’s comment about Vanguard ETFs. They’re a very snug fit to their index after costs.
3. Choose an ETF that tracks a total return index.
4. Choose a physical ETF that doesn’t use swaps or choose a provider that doesn’t do security lending.
Do IT’s lend securities, btw? I’m genuinely asking because I don’t know.
@ Hamzah – On remaining uninvested, the evidence is against: http://monevator.com/lump-sum-investing-versus-drip-feeding/
@ TA: I’m curious, have you bought a momentum ETF yet? And if so, which one?
Hi DonF,
After a lot of research I concluded that a multi-factor ETF best served my needs. I’ve invested in IFSW and written about it here:
http://monevator.com/building-a-multi-factor-portfolio-with-ishares-factorselect-msci-world-etf/
It’s supposed to incorporate momentum, value, size and quality which is the balance I’m looking for. Reputedly multi-factor is more efficient than investing in multiple single-factor funds (e.g. it avoids the cost of selling out of value and buying momentum as stocks migrate).
Hi Accumulator,
Thank you for the reply.
Yes, I’ve been considering investing in IFSW as well. I’m just too tight for its higher TER as I’d eventually like to have a fair part of my portfolio exposed to factors. And 0.22% is less than half of IFSW’s TER…
Besides, I’ve build up quite a heavy weighting towards value over the last few years (GVAL, IDJV, DGS, SEDY as well as a few individual country ETFs that seemed good value at the time – still waiting for Greece to come good, who knows if that’ll ever happen… 😉
So I was planning to just start regularly putting new money into a momentum ETF rather than having to switch around too much.
Decisions, decisions…
PS: I’d love to read a ‘How to invest in’ article about the various Momentum funds that are floating around. Perhaps even taking into account different asset classes? I really enjoyed the articles about how to invest in the Value and Quality factors.
Love the website, by the way. I’m a dedicated reader although I hardly ever comment.
Thanks for the comment, Don F, am I’m very glad you like the site. While you wait for me to get around to momentum, this is a good piece that may whet your appetite for more momentum: http://awealthofcommonsense.com/2015/07/my-thoughts-on-gary-antonaccis-dual-momentum/
Thanks TA. Very interesting, indeed. It almost seems too good (simple) to be true.
Second @Don F #18.
@Time like infinity
As requested in the comment thread on @TA’s latest Commodity piece, here’s an outline of what I am doing for my Momentum ETF strategy, which is currently ~ 37.5% of my SIPP/ISA portfolio.
Since the events of March 2020, I started to adjust my portfolio based on ‘macro’ calls; e.g. I got lucky with swapping my bond ETFs for the BCOG commodity futures ETF from mid-2021, which did exceptionally well until my November 2022, which is when I fully implemented my Momentum ETF strategy which told me to sell it.
I’ve primarily based my technique as laid out by Andreas Clenow in this book ‘Stocks on the Move’.
His technique uses the S&P 500 stock universe, and he recommends having somewhere between 20-30 stocks at one time.
But I am using ETFs listed in the LSE that my broker (HL) offers. I actually started with all ~ 650 of these, and am whittling down as I go along, e.g. if they are too volatile, have large spreads, low AUM, have SRI / ESG / Sustainability / Equality in the title, are distributing (prefer Acc), and I also pick generally pick the largest AUM & tightest spread when there are repeat offerings from different asset managers. So far I am down to ~ 450 ETFs, and imagine I’ll finally have it down to ~ 150.
I think the style I am using is also referred to as Dual Momentum (coined by Gary Antonacci I believe) since I rank by ABSOLUTE momentum, and also compare RELATIVELY to my two benchmarks (FTSE All-World and GBP money market fund).
I taken much input from various articles and papers. E.g. Seeking Alpha has quite few, where people have also done back-testing (I haven’t made the effort to do back-testing myself). I haven’t found anyone describing doing this with LSE listed ETFs though.
• So, as per Clenow, I rank all the ETFs using R-sqr adjusted Exponential linear regression. I use look back periods of 90 (trading) days and 45 days, and then average the two results – the result approximates to an extrapolated % growth per year
• I choose 6-8 ETFs base on those results (currently on 6, and will try to keep it at that)
• I will pass over higher ranked ETFs if they are too similar to ones I already have (e.g. US IT ETF very similar constituents to a Nasdaq ETF at the moment)
• I am using a volatility and max gap filter, so for example, my algo won’t allow to me by a crypto proxy ETF (e.g, check DAGB!)
• I am currently position sizing according volatility (I end up with allocation in my strategy somewhere between 10-25% for each ETF)
• I DO include hedged ETFs. My best performing for the last couple of months have been Nikkei & Nasdaq hedged ETFs
• The exit rules are the most difficult – I’ll check everyday if any of them (as per above described momentum calc.) have fallen below my FSTE All-world or money market fund benchmarks. And then weekly if any are not at least 10% above my two bench marks – then sell if the aforementioned are not met
Currently since Nov (when I fully implemented) I am doing ~2.4% worse than FTSE All-world, and almost exactly the same as a money market fund! (all expenses are taken into account). My underperformance so far has been due to breaking my exit rules because of emotional attachment to miner and precious metal ETFs…. and also suffered nasty whipsawing from those on a couple of occasions.
So, my strategy is evolving. Some things I am trying to iron out:
• My exit rules – should they be same as entry rules or perhaps over a longer period to lessen whipsawing?
• Should I be using single commodity future ETFs? (many of them have horrible spreads) E.g. COCO wasn’t in my ETF list in November… but just look at it now! I would be ahead of my benchmark if I’d included it at that point..
• Should I include a few short ETFs in my list? I do currently have S&P500, DaX, FTSE100 short ETFs, although they of course haven’t been selected by my algo as of yet
• When there is sharp downturn in the stock market, will I be able to get out at an appropriate speed, and get back in at a lower prices ? Most backtesting examples I’ve seen, do seem to suffer less severe dips than 100% stocks or even 60/40 portfolios
• Tighter volatility rules to reduce the whipsawing loss… ?
Hope this post isn’t too long. I need to find a good discussion forum on this with people who are implementing using LSE listed ETFs. Please let me know what you doing and if you know of such forums..
Thank you ever so much @Alegernond. This is fascinating stuff and very much appreciated.
Monevator had 8 factor investing articles in October to December 2014 but, after that feast, there’s been a bit of a famine, relatively speaking.
I’ve been too afraid of concentration risks, p-hacking, overfitting, out of sample issues, replication failure and logistics of implementing to actually pull the trigger, so far, on starting up a DIY concentrated momentum strategy myself, but have been reading into momentum since 2013.
I’d recommend both Gary Antonacci’s 2015 book “Dual Momentum” and Wesley Gray’s and Jack Vogel’s well researched 2016 work “Quantitative Momentum”. There’s a useful reading list (originally from AQR) provided by FundExpert (not behind their paywall) at: https://www.fundexpert.co.uk/fundresearch/blogs/view,the-history-of-momentum-investing-two-centuries-of-pedigree_1803.htm
SSRN and ArXiv also give free access to many of the academic research papers into momentum.
Like you, I also found some of the stuff available on Seeking Alpha useful, but I’ve only read non-pay walled articles there. I don’t know a forum yet for LSE listed ETF momentum specifically, but I do recall the 7 Circles investing blog tried something in 2016 in a variation of the now famed FE “Bonkers” momentum approach (which just chooses the one single OEIC out of the whole universe of funds which has the highest 6 months’ return and then holds it for 6 months and repeats). 7 Circles were trying to make Bonkers less bonkers with some shares and a ETF, also selected by momentum. They then pay walled most of the series, but the first part is still free to access:
https://www.google.com/amp/s/the7circles.uk/the-bonkers-portfolio/amp/
I wondered, perhaps, if on Seeking Alpha you’d spotted Toma Hentea’s Adaptive Momentum Investing system launched in 2021 and, if so, then what you thought of it? On the one hand, with hindsight’s many very obvious benefits, it uses a backtest that seemingly gives the very best of all possible worlds; e.g., for a 6 stock strategy: it uses US large cap dual listed on Nasdaq 100 and S&P 500, with the 2016-21 period for the backtest, a 75 trading days’ lookback period using relative momentum, and something like a 3 months’ hold to review period; all of which are then subject to a trend following switching mechanism to either a risk off regime of differing duration T-Bill ETFs (also selected by relative momentum) or to a risk neutral regime mix of US consumer staples and healthcare ETFs. One the other hand, the claimed results for the backtest look impressive, at least superficially (apparently 68% p.a.). The big question, as always with factors, is whether claimed backtest performance (or anything approaching it) will actually ever repeat again, and – if so – then to what extent and when? I am extremely, highly doubtful of such repeatability myself, but I must also confess to being somewhat awestruck by the backtest results here, despite my very great and persisting skepticism as to whether they can or will be replicated.
There’s also some practical questions about whether the strategy can actually be implemented (without ongoing FX exchange frictions) by a UK based retail investor, to which the answer maybe “partly”, as one can get a multi-currency SIPP through ii but, because of the PRIIS regulations, a UK investor can’t then go and buy the US exchange listed ETF elements, so they’d instead have to presumably sit in $ cash for the risk off regime and then substitute something like BRK.A for consumer staples/ healthcare ETFs in the risk neutral regime.
@ Algernond – Thank you for taking the time to post your system plus links to your sources. Forgive the questions, but why do you track so many positions? And how long does it take you per day to manage?
Incidentally, I think this is worth reading for anyone interested in momentum.
https://earlyretirementnow.com/2018/04/25/market-timing-and-risk-management-part-2-momentum/
@ Time like infinity – 68% per year!? That’s madness. Gotta be backfitted.
I mostly stopped writing about the risk factors because the bulk of the Monevator readership didn’t seem that interested. I still invest in a multi-factor ETF and small value. Can’t say I’ve enjoyed the risk factor premium yet.
@TA: it must be a survivorship artifact. Test 100 models each with 100 variables then, out of 10,000 permutations, 1 likely shows 4 Sigma outperformance.
If 68% p.a. were sustained into the future then it would be higher than Renaissance Technologies’ Medallion Fund gross of fees (63.3% p.a. 1988-2018, circa half that after fees). Even ZXSpectum48k’s former colleagues under Michael Platt over at BlueCrest Capital Management will surely struggle to get anywhere close to that in the longer term.
But FOMO has funny effects. When I see a backtest like that give 68% p.a. for 6 years, then I just can’t help but to be captivated by it. I mean, it does sound like a strategy that one could just about have imagined someone coming up with in 2016 when mega cap, profitable, US tech had rallied hard off its 2009-12 lows.
And the strategy author is 100% bona fide. A retired Professor Emeritus of Electrical & Computer Science. I have no doubt at all that the backtest produced each and all of the results which it says that it did within the given selection of its parameters and variables.
Moreover the fact the strategy creator is from an academic & sciences background is a big positive for me, as I give a higher credence to that professional pedigree than to someone from finance.
But the reality check here is that this can’t be a random strategy choice. It must be what happened to have worked best out of many other strategies that did not.
That provides no logical basis per se, at least that I can see presently, for believing that it will perform very much better than any other similarly concentrated trend following momentum strategy in the future.
But I still can’t help being captivated and then wondering, “what if ..just what if….”
Hi @Time like infinity
I only discovered Momentum & Trend Following relatively recently (last couple of years) from ‘FinTwit’, and then from listening to the ‘Top Traders Unplugged’ podcast.
I have also read Gary Antonacci’s book, but I found for actual implementation, Clenow’s book has more practical detail that was easy to follow. “Quantitative Momentum” – is on my list; I’ll have to make time for it.
I decided to take a risk and implement in my actual accounts last year because I found that I wasn’t motivated to trial it properly running it only in a paper account. I don’t think the risk has been too high, since many of the examples of people implementing seem to show it could work quite nicely, albeit not with LSE listed ETFs.
Thank you for the other sources listed. I will look into them (I don’t pay for Seeking Alpha at the moment, although I am a bit naughty regarding accessing their articles sometimes).
Yes, I have read some Tom Hentea articles. I notice he has looked at using leveraged ETFs also, which seems scary. They can move against you quickly, and have horribly large spreads.
I wouldn’t call what I am doing Adaptive at the moment; it looks quite involved what he is doing to define risk-on / risk-off.
I have been using Clenow’s rather simple rule of ‘don’t buy any new Stock (ETFs for me) if the Market (FSTE All-World in my case) is below the 200-day MA’. And since I have many Commodity / Bond / Money Market ETFs available me, I will have plenty of choice of what to buy when not being able to get new stock ETFs.
Also of course, I can’t get all of the sector ETFs in the UK, which is partly why I decided to start with the entire UK ETF universe and gradually whittle it down. And anyway, if some obscure sounding ETF has high liquidity/AUM & tight spread (e.g. RBTX) and is showing the best momentum, then I’m not sure why I would want to exclude it.
Hello @TA
I’m not tracking many positions. Currently just trying to keep it to 6 ETFs. Once per week (minimum) I’ll pull a year’s worth of end of day data from my entire ETF list (currently an R script accessing Yahoo Finance), but will just get it to display the top 30, and then I can compare those to my current 6 positions.
Daily, I’ll check if any adverse moves for my 6 positions using free charting tool (Trading View), and will take action if any of them have worse momentum than my benchmarks or have dropped below 100-day MA.
For swapping in better momentum ETFs, I will consider that once per week (rules outlined in my original post). The most ETFs that I’ve exited in one week has been 2, and it has averaged about ~ 2x per month in my strategy’s short history so far.
I had read that ERN Momentum article before. It’s good.
I have become quite active since March 2020. Currently I am thinking doing this as a side-hustle is more beneficial that other options for me. I had hoped to FIRE already, but may postpone for up to another couple of years. When I do, spending more time on Momentum, Trend Following, Value Investing is what I plan. It’s pretty interesting, as Finance is linked to so much that is going on in the world.. sends you down all kinds of fascinating rabbit holes. I read a book on Coffee Trading recently, just because I had noted that the Robusta & Arabica prices weren’t trending together!
Many thanks @Algernond.
On Hentea’s use of Leveraged ETFs in one of his portfolios; I think Leverage ETFs are a disaster and should avoided in all circumstances. Do not touch them even if wearing a Hazmat suit and using a bargepole attached to another bargepole. The Evidence Based Investor site did a good take down of Leveraged ETFs here:
https://www.evidenceinvestor.com/leveraged-etfs-another-product-for-the-junk-pile/
And here:
https://www.evidenceinvestor.com/leveraged-etfs-no-thanks/
And investopedia broke them down here:
https://www.investopedia.com/articles/investing/121515/why-3x-etfs-are-riskier-you-think.asp
It’s hard to believe it, but the finance industry has managed to create a category of ‘product’ in the form of the Leveraged ETF that, if held for long enough, is essentially guaranteed to lose money even if the index tracked moves in the right direction for the investor.
More research on UK momentum ETFs has revealed what may be issue unique to factor.
Factor ETFs genetically suffer a definition problem: e.g. is value on basis of P/S, P/E, P/FCF or EV/EBITA; is profitability GAAP or some other measure; is size based on market cap or turnover; is momentum lookback over 3, 6 or 12 months and do you hold for 3, 6, 9 or 12 months (and then, on lookback, do you exclude the last month’s return)?
But momentum ETFs (and the indices they track) seem to suffer a unique issue all of their own, namely a ‘homeopathy’ issue. Instead of seeking out a cohort sized for the maximum risk adjusted momentum with sufficient holdings to capture most diversification benefits (which, at largest, would give the top decile by momentum, but with 10,000 large caps listed globally it could be as small as the top percentile – i.e. 100 holdings); actually available ETFs tracking momentum indices available to UK retail investors seem to only invest in the 30% of stocks by highest momentum. Frankly, that approach is a bit hopeless. Top 30% is not going to effectively capture the momentum factor, if it continues to persist. Same flaw as for homeopathic treatments, which proceed on logically flawed & counterfactual premise that diluting ingredients increases efficacy.
Hello @TLI.
Not sure if you get notified when new comments on this post.. But I felt I should update you that I’ve ended my Momentum ETF strategy.
Not because it wasn’t working, but because it has been too onerous to keep it up (still doing my PAYE job at the same time).
I ran it for just over a year, and when I stopped it at the end of November, I was down compared to my two benchmarks (ACWI & CSH2). My final performance was +0.7% for the year, which is around the same as the various Trend Following UCITS funds have done for the same period, and it was trending the same as those throughout the year, so at least I feel I was doing it approximately right. It’s been a good learning experience.
Will now concentrate on paper accounts in spreadbetting to implement ‘proper’ trend following, so I’m ready to implement for real again when my pension lump sum become available (or maybe a bit before that).
Thank you ever so much for the update @Algernond. Very much appreciated.
The time taken for, and the timing of the execution of, the strategy implementation have been my biggest two reservations about systematic, rules based, quantitative approaches.
What would happen if I was on holiday and missed the risk on to risk off (or risk off to risk on) signal? Or what if there was no coverage for mobile internet to log in and do the trade? Or what if I suddenly had to pull an all nighter at work to meet a last minute, no notice, client deadline, and so missed the signal entirely. Ecetra.
I think it’s likely that some quantitative strategies will work well going forward, but if I can’t pay someone to do the actual execution of them (and, if the option existed, which unfortunately it doesn’t, then I would be willing to pay 2 and 20 for someone to run the very best ideas within a tax free account), then I doubt that I’ll consistently and reliably be able to get the implementation correct every time, year in and year out.
I also think that the effort involved in manual implementation of quantitative approaches and tactical asset allocation means that there’s got to be one heck of an excess return expectation in there in order for the strategy to really be worth attempting in real time for years at a time; i.e. something perhaps a bit like the Dec 2018 TrendXplorer ‘Exploring Smart Leverage: Defensive Asset Allocation on Steroids’ strategy, which uses some leverage and 1, 3, 6 and 12 monthly weighted momentum for the “canary in the coalmine” signal assets (and which achieved a 23.78% p.a. CAGR for the whole backtest period 31st Dec 1970 to 31st Oct 2018). But there’s always the risk of data overfitting with any strategy, even when rigerous in and out of sample testing has been done. And all such strategies fail if funds pile into them if course.
Hi @TLI.
It’s interesting. I don’t think the absolute timing is too too critical (e.g. if close a position one day late). I’m reading Andreas Clenow 2nd edition of ‘Following the Trend’ at the moment (backtested strategies up to the end of 2022). He’s making the point over-and-over, that the exact rules aren’t having the 1st order impact on performance, and that rather it’s the discipline of sticking to the rules without over-riding them (I was guilty of that in what I’ve been doing over the last year).
The main disadvantages of what I was doing with ETFs as far as I see were (in addition to be onerous):
– difficult to target overall volatility / position sizing when using just a few ETFs. I know there are some lleveraged ETFs, but it’s still very clunky
– Not very many short instruments available
– trading cost is prohibitive, there are some quite nasty spreads (e.g. check this Wisdom Tree individual commodity ones, the short ETFs)
With Futures/Forwards (likely to be via spreadbetting for me) you can accurately dial up your desired volatility level, and go long-short in any instrument. Plus of course can really get proper diversification by using many positions across the asset classes simultaneously.
I will set-up my paper account trial using trailing stop losses, so it won’t require daily checking. E.g. I can log in a couple of times a week, and if any positions automatically closed, then open new ones.
Am keen to have a system I can implement in < 2 years time; I feel will be quite an engaging exercise for when not working full time anymore. And I can’t think of any other tax-free thing to put my pension lump sum into outside of ISAs, except for spreadbetting!?
(Thanks form the Dec 2018 TrendXplorer article)
“And all such strategies fail if funds pile into them of course.” @Algernond: One thing which occurred to me after typing those words is that it may not be true of straightforward leverage on market beta. Whilst trend/momentum might be used to target lower volatility regimes to try to avoid the constant volatility trap/volatility drag (as in Michael Gayed’s 200 Day SMA Leverage Rotation Strategy in his 2016 paper “Leverage for the Long Run – A Systematic Approach to Managing Risk and Magnifying Returns in Stocks” on SSRN), it’s not actually looking to exploit an anomaly per se. As such, leveraged beta might not suffer from performance deterioration due to the anomaly/factor being over discovered and over exploited. The problem with leverage though is it’s great on the way up but a killer on the way down. It has to be actively managed when using 2x or 3x LETFs on equity indices and can only ever be considered for a small part (<10%) of any portfolio's starting capital due to the very real and obvious risk of near total wipeout. I’d draw a distinction between those products though and something like WTEF which is 1.5x leverage a 60/40 where only the lower volatility intermediate bonds are levered and the stocks aren’t. That’s much less dangerous. Of course, even modest leverage on low vol assets carries extra risks. But it will be exponentially less hazard than higher levels of leverage on equities.
Interesting conclusions from a 2012 paper on Trend Following (momentum being a form of trend following), Stop Losses, and the Frequency of Trading, using the example of the S&P 500 (by Andrew Clare, James Seaton, Peter N. Smith and Stephen Thomas of Cass Business School & the University of York), namely that:
i) there is no advantage in trading daily rather than monthly;
ii) there is no value in stop loss rules;
iii) ‘whipsawing’ is not a problem, provided technical signals are of reasonable length (not too short);
iv) there is no advantage in complicated trend following rules versus simple rules; and,
iv) trend following rules give superior risk-adjusted returns relative to fundamental financial metrics.