Index | Recent Threads | Who's Online | Search

Posts: 82    Pages: 9    Prev 1 2 3 4 5 6 7 8 9 Next
Last Post
New Thread
This topic has been viewed 3795 times and has 81 replies
ustonapc
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

Scott,

Just to let you know, I just closed and took profit on my bitcoin position around 53500 as the weekly Parabolic SAR just issued a sell signal on bitcoin.

Regards
James

Dec 3, 2021 2:28:47 PM       
Edit 2 times, last edit by ustonapc at Dec 3, 2021 6:09:52 PM
sgmd01
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

James,

Thank you so much for the links. Great job on avoiding the weekend crypto crash. The cleo.one site has potential. My backtests on their site showed the limitations of technical trading when using super volatile asset classes (i.e. they sucked) but I may create a binance account to test further.

Wisdom tree is creating a crypto index. Here's the fact sheet: https://ritholtzwealth.com/wp-content/uploads...ndex-Factsheet_OnRamp.pdf
Many of these coins are significantly more volatile than Bitcoin.

BTW I think SPBC ETF is the best vehicle for accessing Bitcoin if you can't buy the coins directly. It's a combination of 10% GBTC and 90% SPY but without the 6 month lockup of GBTC. GBTC has been offered at a discount between 14-19% for several months now which I think is a better choice that the Bitcoin futures like BITO that will drag behind Bitcoin in profit due to contagio. Unfortunately SPBC is not on P123

Scott

Dec 9, 2021 11:12:54 PM       
ustonapc
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

Scott,

I use Binance to trade bitcoin directly instead of buying GBTC. As Florian mentioned earlier, Grayscale charges a 2% p.a. fee on their trust and the discount to NAV on GBTC has not been disappeared even when bitcoin traded at 69,000 and it will likely remain unless it is later converted to an ETF with the approval by SEC. There is also no 6 months lockup on GBTC, you can buy/sell GBTC anytime with your brokerage account. For BITO, there is around a 15% p.a. "bleed" based on futures contango roll.

The trends of "Alt coins" are less predictable by technical analysis comparing to bitcoin since there is no CME futures (which has lots of technical trading). Alt coins (such as like Solana, Avalanche, Sandbox etc..) moves mostly based on news/sentiments in the market.

I have heard of SPBC but it is better to manage your bitcoin exposure actively in Binance (rather than passively at 10%) and without paying any management fee on SPBC (0.5% p.a.).

It is now a better time to do more research and backtesting (with cleo.one) rather than buying bitcoin due to the weekly sell signal that should remain at least for a few weeks. (I will let you know when I am buying again).

Regards
James

Dec 10, 2021 12:27:45 AM       
Edit 2 times, last edit by ustonapc at Dec 10, 2021 6:04:36 AM
sgmd01
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

Hi James,

Thank you for the good points!

I noticed that P123 now has some grayscale products available for stock simulations as of the beginning of November and a greater number available in books. For ex GBTC and ETHE can be used in sims and books but GDLC (grayscales index) can only be used in a book. One should also be able to backtest this on tradestation as well as cleo.one.

I've read papers on manipulation of the crypto market (Ex: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3639431). From the paper: "We can say with near 100% confidence that bitcoin’s price has been fraudulently manipulated at some point in its lifespan since 2010. We can say with 95% confidence that bitcoin was manipulated in 2013; 95% confidence that bitcoin was manipulated in 2018; and 98% confidence that bitcoin was manipulated in 2019". This is likely far more prevalent with the alt coins that have a smaller market cap particularly since they're available to trade 24/7 so someone can time a thin liquidity period to push the price. In addition they're "questionably" classified as a security be the SEC so they may of may not have legal action against them for this unethical action.

I will look at technical indicators to minimize drawdowns. TA may be self reinforcing with the crypto that are traded through futures. There is considerable timing risk in Bitcoin (and even more in the rest of the alt coins) due to their volatility. For example if one invested 10k in Bitcoin in 1/2015 they would have 246 000 if the ten best days of the whole sample are removed, 11.6 million if the ten worst days of the whole sample are removed, and 1850 if the best day of the month is removed every month. Therefore it may be a good idea to use several indicators simultaneously to reduce timing risk. BTW Bitcoin may be forming a double top. Have you noticed if this leads or is correlated with the rest of the crypto market or do the coins move independently (This paper suggests that Bitcoin leads downturns: https://www.sciencedirect.com/science/article/abs/pii/S1544612319310311, "In the short-run, a decrease in Bitcoin price has greater effect than an increase on the price of all altcoins") ?

As far as trading in futures I noted that it has preferable tax treatment relative to buying the crypto directly (one pays about 10% less in taxes per trade in the US). I understand that the futures contract can considerably deviate from the spot price (I've never traded in futures before). Is there anything else that I should consider if I use futures?

Scott

Dec 12, 2021 10:13:34 AM       
Edit 1 times, last edit by sgmd01 at Dec 12, 2021 10:26:33 AM
ustonapc
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

Scott,

I also believe that the crypto market is heavily influenced (manipulated) by whales in the market. Like what happened in the recent bitcoin crash to the 42,000 level, 15000 bitcoin was dumped in the market within 1 minute. These whales usually trade based on technical breakouts which is why we should use trending following indicator (like moving average or parabolic SAR) to reduce drawdowns.

The correlation between bitcoin and altcoins is high but within different time frames. Ben Cowen thinks that bitcoin/altcoin move in different cycles with bitcoin leading in the beginning of the cycle and altcoins late in the cycle. I won't suggest anyone to place a large position in altcoins since it is very volatile and does not always follow TA. (even a basket of altcoins is more volatile than holding bitcoin).

For example if one invested 10k in Bitcoin in 1/2015 they would have 246 000 if the ten best days of the whole sample are removed, 11.6 million if the ten worst days of the whole sample are removed, and 1850 if the best day of the month is removed every month.


As you have already noticed, it is all about taking out the worst days when trading bitcoin with TA.

Some of the alt coins moves in similar direction, like the layer 2 coins (MATIC, LRC) and GameFi coins (SAND, GALA, MANA) and they are heavily influenced by whales that usually pump and dump based on market news and sentiments.

Regards
James

Dec 12, 2021 1:32:53 PM       
ustonapc
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

Scott,

When you are doing backtesting with bitcoin, don't forget the performance of going short with trend following indicators.

I have highlighted the crypto hedge fund for your reference in the hedge fund performance list. Coincident Capital is a Long/Short Crypto Strategy, primarily trading in BTC and ETH on a swing basis. If you compare its performance vs the benchmark, the fund is up 4 years in a row and over 65% in 2018 when the crypto market tanked. The max drawdown is only 24% during the past 4 years with almost zero correlation with S&P500. Unfortunately, Coincident Capital is currently unavailable for new subscription (probably due to the size limit of its investment strategy).

Regards
James

Attachment Coincident Capital 1.png (45637 bytes) (Download count: 45)


Attachment Coincident Capital 2.png (75083 bytes) (Download count: 44)


Attachment Coincident Capital 3.png (67951 bytes) (Download count: 43)


Attachment HFR Cryptocurrency Index Performance History.png (83295 bytes) (Download count: 44)


Dec 13, 2021 8:24:39 AM       
Edit 2 times, last edit by ustonapc at Dec 13, 2021 10:56:33 AM
sgmd01
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

James,

Thank you. That fund has excellent performance and I would invest if it was open-lol. Where do you get this fund information?

Crypto can have some prolonged steep drawdowns (5 months to a year vs equities at 3-5 months if one counts peak to trough) so I can see why they would also go short. The question is will the length of the drawdowns in crypto persist as the asset class gains greater acceptance (rhetorical question)? Crypto also can have near vertical price increases when coming off the bottom so there's a lot of timing risk. One cannot backtest short strategies on cleo.one. I'm not sure about tradestation yet but will update you when I find out. We could test strategies here with ethe and gbtc once P123 expands the data set.

Scott

Dec 13, 2021 12:35:00 PM       
ustonapc
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

Scott,

I obtained the hedge fund information from Barclay Hedge Database and the performance table from Nilsson Hedge Database.

This is the backtested performance on bitcoin for the past 4 years when I use Binance to manually test weekly parabolic SAR and only long/short when ADX (7) weekly is also >35

Bitcoin SAR (0.02,0.2)
2018 45%
2019 127%
2020 530%
2021 38%
4 years Annualized 131.29%

https://www.binance.com/en/trade/BTC_USDT?layout=pro

It will be interesting to find out whether you can backtest short positions with tradestation. Pls update me when you find out more.
According to Yuval, adding GBTC, ETHE and other grayscale trust for sims is working in process.

Regards
James

Dec 13, 2021 1:00:25 PM       
Edit 2 times, last edit by ustonapc at Dec 13, 2021 1:01:52 PM
sgmd01
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

James,

That's excellent performance. Thank you for the link and information. Here's a link to an ARK paper on valuing Bitcoin: https://ark-invest.com/articles/analyst-research/valuing-bitcoin/

Scott

Dec 13, 2021 1:43:35 PM       
ustonapc
Re: Bitcoin (PlanB : Stock-to-Flow model and Ben Cowen : Algorithmic Regression Rainbow)

Scott,

I don't buy ARK research anymore, just look at the performance of their funds this year. All of their funds (including ARKW which has an allocation to GBTC) except one in the whole ARK family are down this year. (i.e. underperforming S&P 500 by 20%-50%)

Regards
James

Attachment ARK.png (80426 bytes) (Download count: 32)


Dec 13, 2021 1:58:44 PM       
Edit 2 times, last edit by ustonapc at Dec 13, 2021 2:22:36 PM
Posts: 82    Pages: 9    Prev 1 2 3 4 5 6 7 8 9 Next
 Last Post