I don't understand. I know what [hf]odl mean. Can anyone tell me what these charts and conclusions are trying to communicate here?
edit: I mistakenly assumed the HN would the understand the regex of "[hf]odl" was just a clever way of saying hodl/fodl. No need for more comments explaining those terms.
I think its comparing how the algorithm did versus just blindly HODLing your coins. So HODLmeter:XX% is the percentage the algorithm increased the value of its holdings over the described period. The HODLing is the percentage the market increase by over the same period. Thus, if you had HODLed for the same period for ETH, you'd have lost a significant amount in the value of your holdings, but the algorithm made money.
HODL is a memetic misspelling of "hold", meaning to hold on to an investment. FODL is a play on HODL refering to folding in poker, meaning to give up on a gamble and take a loss in order to avoid a larger loss later. Usually by trading for a stabler coin like Tether.
The big HODL/FODL means if you should hold/fold right now. Hodling percentage means your return over the period when holding, and HODLometer percentage is your gains when using the algorithm.
Another day another trading algoritm. This is somewhat similar to the n+1 problem with standards. If you had n algorithms and come up with another one claiming to outperform the rest, all you have now is n+1 trading algorithms.
If there was an algorithm so good that it could influence the market, then the market would self correct and the algorithm proven useless.
Would love more details on how the algorithm works, and your thought process behind it (e.g. what other indicators you've looked at, why). You don't have to disclose your secret sauce, but it would give me a lot more trust.
Algorithm have been found on TradingView, and tested among many.
Cryptocurrencies are especially volatile that's why I thought I would need to use ATR. I have also been playing with trailing stop and take profits, so found this algorithm to combine the two.
The parameters have been manually tested, but it would make sense to make run a simulation to get the best parameters for a coin.
The whole site was created in about 6 hours, so don't judge too hard.
Not necessarily, depending where you are geographically and what counts as taxable event.
Also it's not a lot of trades for a year, after all, so not a lot in commissions either.
Sure it shows the algorithm beats a hold strategy in 8/10 instances...
...but the "hodlometer.com" domain was registered 3 days ago.
It's trivial to develop an algorithm by cherry-picking parameters that will give any desired results for historical data.
But there's vast research on how it's impossible to predict/beat the market like this in reality [1], and this site is obviously designed to prey on the uneducated and naive.
[1] For a good intro, see chapter 5 of "A Random Walk Down Wall Street" which talks about "technical analysis", which is exactly what HODLometer is doing -- https://www.amazon.com/gp/product/0393352242
People who know a thing or two about finance are having the time of their lives watching Bitcoin evangelists genuinely believe in technical analysis and not understand that it's on the level of astrology and tarot cards as far as usefulness in predicting prices. (And that's in a fair market. On Bitcoin exchanges, where it's certain that the exchanges are using every trick in the book-- frontrunning, painting the tape, wash trading etc.-- to stack the deck and extract extra value from your trades, technical analysis is even more useless)
There is a basic pattern (bubble/pop/settle) but for any advantage you have to know better than someone else. There's been a number of images showing graphs scaled and overlaid over other graphs. There's quite possibly good reasons for the correlation in shapes, but it is also quite useless for investment. You might know something will probably happen but you don't know when or at what speed.
Once you go past the patterns that everyone can plainly see, you are looking for patterns in what appears to be noise to everyone else. Accurate future predictions is really the only way to tell if it is anything other than staring at tea leaves.
It's taking all available data from Binance API (Basically all the different cryptocurrencies have been listed at different times, therefore the difference in time ranges)
I can see though how it looks untrustworthy because of that, thanks for feedback!
The philosophy behind HODL is that you need to hold for 5-10 years plus in order for the strategy to work. The 1-year timespan in the graphs on the site are way too short.
If you want to cherry pick a point to "disprove" HODL, just look at the fact that in November 2013 Bitcoin traded at an eye-watering $950+, and then for 2.5 years beat a steady retreat to $400.
At this point, bitcoin can taper down to $3K in value till 2020 and we would not have seen anything new in terms of relative price movement over a fix timespan.
For those who have watched bitcoin for a while, 2018 is just 2014 all over again, except this time a lot more people are paying attention.
Bitcoin's a deflating currency; there's a fixed number that will ever be mined, and a number are lost to forgotten private keys etc, and so the supply will only ever go downwards.
Meanwhile, if its boosters can actually keep it in the news, build things with it, and use it as a currency, demand for it in the long run will continue going up. This is the other unspoken assumption of HODLing, unspoken because most people don't want it explicitly known that they're hyping & propagandizing, but pretty apparent in their actions. This is also not true in the short run; hype cycles like what we saw in Dec 2017 are impossible to sustain forever, which is why the price today is significantly lower than it was then, which is why HODLing only works on long (5-10 year) timespans rather than short ones.
26 comments
[ 1.9 ms ] story [ 91.1 ms ] threadedit: I mistakenly assumed the HN would the understand the regex of "[hf]odl" was just a clever way of saying hodl/fodl. No need for more comments explaining those terms.
https://bitcointalk.org/index.php?topic=375643.0?red
If there was an algorithm so good that it could influence the market, then the market would self correct and the algorithm proven useless.
Backtested returns are not reliable.
Algorithm have been found on TradingView, and tested among many.
Cryptocurrencies are especially volatile that's why I thought I would need to use ATR. I have also been playing with trailing stop and take profits, so found this algorithm to combine the two.
The parameters have been manually tested, but it would make sense to make run a simulation to get the best parameters for a coin.
The whole site was created in about 6 hours, so don't judge too hard.
...but the "hodlometer.com" domain was registered 3 days ago.
It's trivial to develop an algorithm by cherry-picking parameters that will give any desired results for historical data.
But there's vast research on how it's impossible to predict/beat the market like this in reality [1], and this site is obviously designed to prey on the uneducated and naive.
[1] For a good intro, see chapter 5 of "A Random Walk Down Wall Street" which talks about "technical analysis", which is exactly what HODLometer is doing -- https://www.amazon.com/gp/product/0393352242
Once you go past the patterns that everyone can plainly see, you are looking for patterns in what appears to be noise to everyone else. Accurate future predictions is really the only way to tell if it is anything other than staring at tea leaves.
How would you suggest entering short term trade scenarios then?
I sound snarky, but I'm genuinely curious what you suggest instead.
I can see though how it looks untrustworthy because of that, thanks for feedback!
If you want to cherry pick a point to "disprove" HODL, just look at the fact that in November 2013 Bitcoin traded at an eye-watering $950+, and then for 2.5 years beat a steady retreat to $400.
At this point, bitcoin can taper down to $3K in value till 2020 and we would not have seen anything new in terms of relative price movement over a fix timespan.
For those who have watched bitcoin for a while, 2018 is just 2014 all over again, except this time a lot more people are paying attention.
Meanwhile, if its boosters can actually keep it in the news, build things with it, and use it as a currency, demand for it in the long run will continue going up. This is the other unspoken assumption of HODLing, unspoken because most people don't want it explicitly known that they're hyping & propagandizing, but pretty apparent in their actions. This is also not true in the short run; hype cycles like what we saw in Dec 2017 are impossible to sustain forever, which is why the price today is significantly lower than it was then, which is why HODLing only works on long (5-10 year) timespans rather than short ones.
Reduced supply + increased demand = higher price.
USDT? I don't think that is a good advice. Maybe a better one will be to sell BTC in exchange for a HODL coin instead?