Excellent video ..👌🏻👌🏻 cleared my doubts on 12 minus 1

Nice interview ! I'm personnaly using a Dual Momentum model with multiple timeframe in crypto, weekly rebalance. That's work pretty well. Moreover, data is easily available and the trading cost are low. (Binance, FTX) Now, I'm thinking to apply some short mean reverting during negative absolute momentum time to capture some little up. Do you know if there are a lot of hedge fund who apply momentum approch on crypto ? Or just a few ?

@ExcessReturns

2 жыл бұрын

Hey Maximum, we can't say for sure but we would imagine most HFs investing in crypto are likely using some momentum signals and trend following as part of their buy/sell discipline. Paul Tudor Jones was allocating to Bitcoin as an inflation hedge (not sure how that has worked out so far), but historically he also utilized trend following so it would seem logical he also be using TF with Bitcoin. Hope that helps. Thanks for watching Excess Returns.

I was looking forward to this interview but the audio quality was really patchy and many of the ans got trun . It's q fru ting to try to lis n to the con ation. I also felt that many of the question you asked were never actually answered (eg do we ever get a clear description of absolute v relative momentum ? Maybe I need to listen again). Anyway, thanks for trying. A+ for effort :)

@ExcessReturns

2 жыл бұрын

Sorry about the audio quality and we appreciate your support. I don't think Gary gets too specific on the exact metrics he is using in his model but in general absolute momentum could be something like ... does the index have a return above the 10 year Treasury bond over the past 12 months or is the index above or below it's 200 day moving average. Those are ways to measure absolute momentum. Relative momentum is comparing indexes against each other based on some type of momentum indicator, such as relative strength of 12-1 month momentum. Hopefully that helps and gives you some insight on these momentum type indicators.

Hi chaps, excellent content as always. I wonder if you could elaborate on how retail investors could and should implement multiple look back periods? I am content with the 12 month look back period in Gary’s book, however I note some more recent recent has suggested a 6 month period has been superior in recent markets. I hear indeed that Gary utilities multiple look back periods now also… how would I go about implementing this? Perhaps moving averages rather than absolute?

@ExcessReturns

2 жыл бұрын

Hey Josh, it's a good question. Most academic studies show intermediate momentum is best, which is why they use 12 months minus the most recent month. As Gary mentioned, I think he is using multiple now and one could argue that using something like 6 months might be optimal in a world where you have these declines and sharp reversions higher. It's something we'll continue to keep our eyes open for and if we see any new data coming out supporting shorter look-backs we'll keep you posted.

@Huxley350

2 жыл бұрын

@@ExcessReturns it’s certainly interesting, from a very basic 5 year backtest I carried out, 12 months still looked better but that’s probably just because it kept me in the market more often than not… from a practical perspective, how would one “combine” two look back periods?

@ExcessReturns

2 жыл бұрын

@@Huxley350 Say you are using 12 and 6 months, you could take the average of the two and hold the positions that rank highest. The other approach might be to use the 12 months on half of the portfolio and six months on the other half. There are probably other ways to blend, but those are two that come to mind.

@Huxley350

2 жыл бұрын

@@ExcessReturns thank you! I think I was overthinking it… I like the simplicity of allocating 50% to 12 months and 50% to 6 months. Very helpful thank you.

Can someone comment in more detail on his approach to bitcoin? What algorithm specifically is he using for buy and sell indicators?

@ExcessReturns

2 жыл бұрын

A lot of what is in Gary's models is proprietary, but in this paper he states he is using a simple moving average. Given Bitcoin's volatility, it's likely he is using some shorter term moving average in his testing. dualmomentum.net/2021/06/23/bitcoin-a-sensible-approach/

Good stuff, all around in trying to predict gains in the market and finding those stocks and funds that are going to push through the loss of that mean reversion. What do you all think of the Sound Mind Investing approach of the 3M Wtd 12M MOM look back period of adding the 3M, 6M, & 12M periods rebalanced monthly? According to Porfoilio Visualizer, the 25D MOM look back is the most profitable in this COVID environment, which is unkind to the standalone 3M MOM period as it is the inverse of the 25D MOM. I’ll probably delete this post soon and try and make it a proprietary. 😮

@ExcessReturns

Жыл бұрын

Hi Greg, thank you for watching and commenting. Generally speaking, using multiple lookback periods over an intermediate term time horizon (3-12 months) is good as it helps reduce the over-reliance on just one period of momentum. It's interesting the 25D has worked best over the past few years. Most research shows you don't want to use one month momentum due to short term reversals. Thanks again!

@Bargab1

Жыл бұрын

Yes, as I’ve been doing my calculations on both funds and stand alone stocks, there has been a high correlation between that 3M Wtd 12M MOM and the 25D MOM periods as well as an high inverse correlation between the stand alone 3M MOM and the 25D MOM in this COVID environment. However, there are a few instances, mostly in the stocks where the 25D MOM is in the red, or negative, and the 3M Wtd one is in the green. The 25D MOM does push through that 1M reversal, but didn’t Antonochi say that that mean reversion is more 1-2 weeks when you all were asking him about cutting out the the first month?

@Bargab1

Жыл бұрын

By the way, that 3M Wtd 12M MOM does use your recommended multiple look back periods, but it’s all in one formula. Any thoughts on it?

@Bargab1

Жыл бұрын

According to the Relative Momentum model on the Visualizer, the 5D has a 1% gain and the 20D has an 11% gain, but the 10 & 15D MOM show losses of -8.80% and -0.19% respectively, but the 25D MOM sits at 13.67% to the S&Ps 9.48%. This suggests that Antonach’s 1-2 week mean reversion is accurate as opposed to the one month reversion. Any thoughts here?

@Bargab1

Жыл бұрын

All of this is within the past two years of COVID and the trillions of dollars spent by our government and all the other restrictions with the 3M Wtd 12M MOM sitting at a gain of 10.47% to the S&Ps 9.48% over these past two years since Jan 2020. I should calculate the differences over the past year.