When it comes to stocks or crypto, how do you "average down?"

[WorldCitizen99]

Let's say you are bullish on crypto X and take your entry position at a price of say, $1, per coin.
Then the next day it falls by e.g. 5%. Do you buy some more? Do you use technical analysis to decide? Do you decide on how big your tranches will be ahead of time?
What is your method for how much to allocate to your entry position, and then how to average down (or up)?

 

[JohnnyDoe]

If something is worth buying today at price X, it must be even worthier tomorrow at a lower price, if nothing else changed.
I buy crypto every Monday and I don’t check the price.

 

[WorldCitizen99]

If something is worth buying today at price X, it must be even worthier tomorrow at a lower price, if nothing else changed.
I buy crypto every Monday and I don’t check the price.

OK so you are a true DCA man - very reasonable
And I guess Monday is a deliberate choice based on the fact that last bullrun, Sunday was red day practically every week?

 

[JohnnyDoe]

OK so you are a true DCA man - very reasonable

No I don’t do DCA and martingale style strategies, they are very dangerous.

And I guess Monday is a deliberate choice based on the fact that last bullrun, Sunday was red day practically every week?

I receive fiat payments on Mondays, which I immediately use to buy crypto. That’s the only reason why I chose this day.

 

[void]

in general you (should) have a price model and money management strategy - both is always subjective and your job to do
the simplest price model is "I buy for any price" or "I buy anything under price X until day Y"
the simplest money management approach is "I buy for any available funds"

both may seem extreme but perfectly fine if you know what you're doing and why

traders or those who are not just stacking will have more sophisticated probabilistic price model based on vector of inputs and money management strategies that will most likely be some variation of the kelly criterion - in general trying to maximize their utility function

you can find plenty of resources on this topic, and you'll have lots of fun studying

 

[WorldCitizen99]

OK that is a good answer. I will look up some of those resources. I don't have a defined money management strategy other than pick what I think are good quality crypto projects, pick an entry point on a red day, don't FOMO in, and be prepared to hold until the next bullrun and then sell in tranches when I think we are near the top. That's about it.
Also I found that in the past when I traded on shorter time scales, my life was miserable. So I might do the occasional swing trade, I tend to HODL and scale out of the postion after it's run for awhile.

 

[khinkali]

more sophisticated probabilistic price model based on vector of inputs and money management strategies that will most likely be some variation of the kelly criterion

Great answer.

@WorldCitizen99 I think it's worth being really clear on how you separate your risk management from the alpha model.

For risk management you are unlikely to know the probability of a win or the return if you win to anywhere near the accuracy to use something like the Kelly Criterion, so it's probably best just to focus on what you can afford to lose. You can aim for uncorrelated bets to help with risk, but in crypto this is not usually practical unless you're incorporating short trades into your strategy.

I see many people doing mental gymnastics, trying to balance trend and reversion in some way that will create alpha. I don't think that "pick an entry point on a red day" is going to do you much good. Did it move because it's trending and will continue, or after moving is it likely to revert back? To do the work to answer that, you're life could become miserable again.

pick what I think are good quality crypto projects

That is your alpha model. If you value the project based on your rules, then you know what your buy and sell prices are, irrespective of what other people and bots are putting in orderbooks and AMMs from moment to moment.

 

[WorldCitizen99]

Great answer.

@WorldCitizen99 I think it's worth being really clear on how you separate your risk management from the alpha model.

For risk management you are unlikely to know the probability of a win or the return if you win to anywhere near the accuracy to use something like the Kelly Criterion, so it's probably best just to focus on what you can afford to lose. You can aim for uncorrelated bets to help with risk, but in crypto this is not usually practical unless you're incorporating short trades into your strategy.

Yes many of the tokens are correlated with BTC and ETH, but there are moments when the whales sell their big tokens so they can pump up the value of smaller projects and the 2 markets momentarily become quite uncorrelated. But outside of that, the market moves in unison.

>I don't think that "pick an entry point on a red day" is going to do you much good. Did it move because it's trending and will continue, or after moving is it likely to revert back? To do the work to answer that, you're life could become miserable again.

Yes you're right - depending on where one is in the cycle, red days might beget more red days. With crypto, the cyclicality (so far) has been more pronounced than in other markets. It seems to go in 4 yr cycles so far so if you just make sure you're on the "front side of the hill" then the strategy of buying on down days is a reasonable risk management method if one is willing to wait it out and has plenty of spare underwear for the volatility.

That is your alpha model. If you value the project based on your rules, then you know what your buy and sell prices are, irrespective of what other people and bots are putting in orderbooks and AMMs from moment to moment.

Crypto is hard to valuate when a token is first released. Projects spike based on the excitement around a totally novel idea or a solution to an existing snag in the way blockchains function. In the end it might just be vaporware or a whitepaper that dazzles with theory but is short on delivery. Still, crypto has given gains way above traditional markets (and losses too of course)

 

[void]

Crypto is hard to valuate when a token is first released

for a randomly chosen new token or first layer project the probability (my own model if you will) is almost one

which doesn't mean one cannot make good money using the opportunity and shave the idiots

 

[khinkali]

@WorldCitizen99 I agree there are lots of moves, uncorrelated to BTC and ETH prices. The risk management issue for long-only trading is that there are days when many tokens drop together.

It seems to go in 4 yr cycles so far so if you just make sure you're on the "front side of the hill" then the strategy of buying on down days is a reasonable risk management method if one is willing to wait it out and has plenty of spare underwear for the volatility.

If the hill is going up then you might have been better, not waiting for the dip. Same in a down trend where it can cost money, waiting for a brief spike in price to sell. My number one task has been to have the system be able to make a series of neutral to positive yield long trades during down trends without giving up on catching the big upward breakouts. It's easier to do this when making the market, so you can add the spread to your profit.

I won't trade new tokens because there are no data for me to decide on a price. Also I'm trying to get my mean holding time down below ten hours, which is probably different from what you're looking to do.

 

[WorldCitizen99]

for a randomly chosen new token or first layer project the probability (my own model if you will) is almost one

which doesn't mean one cannot make good money using the opportunity and shave the idiots

Can you spell out what you mean a little more...

Do you have a model for buying and selling new tokens that has worked for you?

@WorldCitizen99 I agree there are lots of moves, uncorrelated to BTC and ETH prices. The risk management issue for long-only trading is that there are days when many tokens drop together.


If the hill is going up then you might have been better, not waiting for the dip.

Yes that is always the dilemma I have - the hill goes up in a general sense but there may points where it dips down below your entry point even if that was a few months ago....I tend to make a firm decision that I will take a position of say 50% of the total money I plan to put in and then bc of the volatility, I will often get a better price later on. Crypto (as Im sure yu know very well) is the wildest rollercoaster in the park and red days or weeks of -30-40% happen periodically....
But I wrote this thread to get some advice from people who are more experienced traders than me, not to pretend like I am an expert...so do you think I should put all my capital in, in one shot, right at the beginning?

Same in a down trend where it can cost money, waiting for a brief spike in price to sell. My number one task has been to have the system be able to make a series of neutral to positive yield long trades during down trends without giving up on catching the big upward breakouts.

Wait - do you go long on both sides of the hill, or just on the back side of the hill? I had never thought of going long during the crypto down trend, (which can be like taking an elevator ride after the cord has snapped!) and profiting from the upward rebounds. Did I misunderstand you?

It's easier to do this when making the market, so you can add the spread to your profit.

I won't trade new tokens because there are no data for me to decide on a price. Also I'm trying to get my mean holding time down below ten hours, which is probably different from what you're looking to do.

 

[void]

Can you spell out what you mean a little more...

Do you have a model for buying and selling new tokens that has worked for you?

no
I posted from the phone and I accidentally lost "of failure" from the statement that "probability of failure is almost one" - sorry to disappoint you...