strategy Is there a correlation between values of numeric portfolios?

Interesting concept

I might be the only person looking at them like this.

Number domains are really just a pure investment speculation at this point. They do sell and there are differences in value due to the chinese meaning and "lucky" numbers etc..

Looking at them as just numbers, there has to be a correlation between portfolio values just based on the length/rarity of the domain. Since they are not keywords, their value is only based on their character length when we take out the "lucky" numbers. Number portfolios like 5N and 6N are very different than a 5L or 6L.com

Looking at a domain just based on rarity and length.
5N.com vs 5L.com both are only 5 characters in length.
5N.com is rarer than 5L.com
5N.com has 100,000 combos
5L.com has 11,881,376 combos
5N is over 118x rarer than 5L

Going past 4L.com and 5L.com you need the keyword to make the domain have any real value. That might change in time though. At this point we are not seeing a buyout in random letter strings into the high character count. 5L.com yes but only because 4L.com is gone and value has increased in 4L.com
I'm not really factoring in the pinyin letters, because I'm not sure up to what letter character they matter past 4L.com

Buyouts on the number side are happening up to 7N-9N.com. Yes the "lucky" numbers get bought out first, but then the rest follow even the 4 and zero, which time is showing do sell.

I got way off track with that, but I think as time goes by and sales data come in, we will see a relationship between the value of each number portfolio. Maybe as we go up one character count, we see the value decrease by a factor of 10x or 100x. It will be very interesting to see if this does happen. We need more sales data to support this idea.

Now just sell some

HireDomains said:

Interesting concept

Click to expand...

Interesting concept; and it requires lots of valuations from here to there, from that to this. In other words, 360degress statically valuation. And it's only based on statically estimation.

ZapNano said:

And it's only based on statically estimation.

Click to expand...

From 3N to 4N.com we have sales data for those numbers. Going beyond 4N we have to estimate things more due to the sheer number of domains and the recent buyouts. The sales data has to come in more on 5N and above to prove any theory. 5N and 6N sales have only really taken off in the last 1-2yrs.

HireDomains said:

Now just sell some

Click to expand...

I'm not really a seller. I wish I was, but I tend to hold onto my domains for the long haul. It's getting too hard to find a good domain for a decent price, so when I do, I hold it.

We are still in the beginning of the number domain market, so those are all long-term investments for me.
I still see way more potential in domains than the avg person. If I really need the money or get a really great offer, I would sell, but until then...I'm hanging on for the ride. Good luck to all

Yes I do believe that the 5n-9n market will only grow

My post from October which relates to a number of my posts from last Spring/Summer...

https://www.namepros.com/threads/whats-the-n-scale-now.884132/

Essentially, I agree with the concept.... clearly there is a correlation, but it's a bit hard to pin it down.

My 5 cents on this. I am only speaking of Numbers, since I never got into trading letters. But I traded enough numbers to know how much I regret trading them away. <>

Everyone knows that .nets are about 1/10 of the value of .com (on average). So according to your logic, 3N.net would be worth 4N.com. Even if we factor 4s into both, you know this is incorrect ratio. 4N.coms are selling at about 30-35K while 3N,nets do not cost 30-35K, unless you have absolute premium without 4s. In other words, It seems like 4N.coms are growing out of your scale faster then 3Ns. Now, according to this same logic, 5N.coms should be worth 1/10 of the 4N.com, and once again we would have to be talking about premiums. If you add a 4s in there, it's more like 1/15 - 1/20th. Maybe 4Ns are out of scale, like this because of the timing of when they were caught on fire and resold many times over as China was entering the market. Way before China, we used to trade 3Ns and noone ever thought anything of it.

My main point here is, while there is an obvious correlation - and nice concept btw, it's not linear, due to the rarity of the 3Ns and 4Ns, who seem to shoot out in price faster.(and timing of 4N entering the hot market) I bet 5Ns would catch up at some point and become more linear, and then they would fit better into your idea. Once they do, it would be time for trading 6Ns, and those would take time to catch up and line up on the graph., and so on.

heavend said:

My 5 cents on this. I am only speaking of Numbers, since I never got into trading letters. But I traded enough numbers to know how much I regret trading them away. <>

Everyone knows that .nets are about 1/10 of the value of .com (on average). So according to your logic, 3N.net would be worth 4N.com. Even if we factor 4s into both, you know this is incorrect ratio. 4N.coms are selling at about 30-35K while 3N,nets do not cost 30-35K, unless you have absolute premium without 4s. In other words, It seems like 4N.coms are growing out of your scale faster then 3Ns. Now, according to this same logic, 5N.coms should be worth 1/10 of the 4N.com, and once again we would have to be talking about premiums. If you add a 4s in there, it's more like 1/15 - 1/20th. Maybe 4Ns are out of scale, like this because of the timing of when they were caught on fire and resold many times over as China was entering the market. Way before China, we used to trade 3Ns and noone ever thought anything of it.

My main point here is, while there is an obvious correlation - and nice concept btw, it's not linear, due to the rarity of the 3Ns and 4Ns, who seem to shoot out in price faster.(and timing of 4N entering the hot market) I bet 5Ns would catch up at some point and become more linear, and then they would fit better into your idea. Once they do, it would be time for trading 6Ns, and those would take time to catch up and line up on the graph., and so on.

Click to expand...

I think the failure in your response is trying to bring 3N .net's into this. .nets are not following very proportionally anymore.

Given that there's on 1,000 of them, they don't trade often... but even when they do, they are not 1/10th of recent 3N .com prices.

http://namebio.com/?s=wMzEDNyUzM
489.net 7,150 USD 2016-01-24 SnapNames
045.net 5,001 USD 2016-01-15 GoDaddy
755.net 2,500 USD 2015-11-22 Sedo
238.net 9,130 USD 2015-09-02 Sedo

And with those, I'm sorry, but there's no way on January 24th, 2016, 489.com would be worth a mere $71,000. Back in November, no way was 755.com is worth a mere $25,000. Not even close. Maybe 1/500th based on that sale, or a worse ratio!

With so many other extensions, IMHO, the .net thing has been watered down a good bit and the break down is only in the .net theory of your reply. I'd suggest not bringing that into an analysis of the .com numerics.

Now... As for 4N vs 5N, they do hold up decently well in scale...

4N:
5233.com 51,772 USD 2016-02-03 NameJet
4820.com 13,643 USD 2016-02-03 Sedo
2230.com 24,944 USD 2016-01-28 Sedo
3112.com 36,200 USD 2016-01-27 NameJet
http://namebio.com/?s=AO5EDNyUzM

5N, but you have to be more selective to find comparators simply due to the volume difference:
76929.com 3,200 USD 2016-02-03 NameJet
75729.com 2,500 USD 2016-01-29 Sedo
77152.com 3,287 USD 2016-01-27 4.CN
77325.com 3,193 USD 2016-01-27 4.CN
13713.com 5,900 USD 2016-01-24 NameJet
55130.com 4,500 USD 2016-01-24 GoDaddy
http://namebio.com/?s=gNwIDNyUzM

Granted, you have to accept there are a number of variables... And as the scale moves out, the variables increase. IE, in 5N, you have a lot more variety in patterns than you do in 4N. Thus the variations are going to be more open to interpretation, as will be seller motivations in a more vast inventory.

Well, when it comes to NNN.nets, I have to disagree. See, I sold over 35 of them in the past 2 years, and bought a few. For some reason Chinese like .nets as well, even through they do price them differently of course. Now, .orgs on the other hand were watered down with the rest of the new extensions and lowered in price, but even they have been trading really nicely.

In fact my 2-3 most recent sales of NNN.nets were in the $15,000 - $17,000 range. I am not sure how I missed the 755.net for $2,500 - must of been asleep at the wheel. - but it's a steal.

One thing that this proportion valuations miss is memorability.

The proportion ratio does work for LLL and LLLL because our brain can still work with 4L acronyms. We start losing it with 5L unless there is a big logic, pattern or pronounceability. That is why Random chip 5L is not 1/20 of random chip 4L, otherwise they'd be trading at $100 each, while, I am sure, you can still hand reg many of those. At the same time, 4L is about 1/20 of 3L, so it does translate well and 1/20 ratio works fine with 2L as well.

With Numerics, with each added number the memorability suffers considerably. We would forever remember NN.com we'd see in an ad. Many here would have hard time to remember exactly all their NNNNNs, if they have few. And even ignoring all that lucky number crap (sorry, personal opinion), adding additional N reduces value by more than 1/10 and each additional N washes it even further.

So if most NNN.com is worth $0.5-1.0 M, most NNNN.com is worth 20K-50K, about 1/20
Most 5N would be $1K-3K, another 1/20.
Most 6N would be $40-150, further 1/20.
And that is why most of 7N are still reg fee or below.

So, yes, it seems 1/20 fits better with real life pricing than 1/10 or 1/8 (Chinese lucky stuff).

Rather than try to rationalize a category, perhaps rationalizing an individualized and comprable set would make it easier to see correlations.

For example... Appraise/price these:

98.com
989.com
9898.com
98989.com
989898.com
9898989.com
98989898.com

and while this is a great example, we should also appraise and price these next to the ones above... (on the low podium of each), so then we can take an average price and see the correlations:
04.com
064.com
0164.com
04164.com
016204.com

I think with low end of it we can stop here, and the others will not coast above reg. And then the average would show real correlation. Also, if we go back in time and appraise it as we were a year ago, two years ago, three years ago, we would potentially see a useful trend.

21x20 said:

Rather than try to rationalize a category, perhaps rationalizing an individualized and comprable set would make it easier to see correlations.

For example... Appraise/price these:

98.com
989.com
9898.com
98989.com
989898.com
9898989.com
98989898.com

Click to expand...

This is good example. In this set the memorability does not suffer much, as you add another "shortcut" in your brain (last even number before hundred, then repeat the first digit). So you could argue that 989.com should be much closer in value to 98.com than 1/20. Perhaps in 1/5-1/10 range.

But 098, 198, 298 etc. is not as easy as 989, so it still would be 1/20 of value minus some discount from 098 for "meaningless" zero in front. 980 should not get any discount though.

Recons.Com said:

One thing that this proportion valuations miss is memorability.

The proportion ratio does work for LLL and LLLL because our brain can still work with 4L acronyms. We start losing it with 5L unless there is a big logic, pattern or pronounceability. That is why Random chip 5L is not 1/20 of random chip 4L, otherwise they'd be trading at $100 each, while, I am sure, you can still hand reg many of those. At the same time, 4L is about 1/20 of 3L, so it does translate well and 1/20 ratio works fine with 2L as well.

With Numerics, with each added number the memorability suffers considerably. We would forever remember NN.com we'd see in an ad. Many here would have hard time to remember exactly all their NNNNNs, if they have few. And even ignoring all that lucky number crap (sorry, personal opinion), adding additional N reduces value by more than 1/10 and each additional N washes it even further.

So if most NNN.com is worth $0.5-1.0 M, most NNNN.com is worth 20K-50K, about 1/20
Most 5N would be $1K-3K, another 1/20.
Most 6N would be $40-150, further 1/20.
And that is why most of 7N are still reg fee or below.

So, yes, it seems 1/20 fits better with real life pricing than 1/10 or 1/8 (Chinese lucky stuff).

Click to expand...

Reconds, likes to write long stuff

Recons.Com said:

This is good example. In this set the memorability does not suffer much, as you add another "shortcut" in your brain (last even number before hundred, then repeat the first digit). So you could argue that 989.com should be much closer in value to 98.com than 1/20. Perhaps in 1/5-1/10 range.

But 098, 198, 298 etc. is not as easy as 989, so it still would be 1/20 of value minus some discount from 098 for "meaningless" zero in front. 980 should not get any discount though.

Click to expand...

Part of what I'm getting at is that you can't easily compare apples to oranges and expect the same results.
IMHO, but this set is apples to oranges:

04.com
064.com
0164.com
04164.com
016204.com

... Because you further devalue the extended names by mixing up the numbers to be more randomized. You are not just isolating the value based on length. You are also devaluing based on the randomization of the numbers in addition to the length extending outward. That's two variables instead of just one.

21x20 said:

Part of what I'm getting at is that you can't easily compare apples to oranges and expect the same results.
IMHO, but this set is apples to oranges:

04.com
064.com
0164.com
04164.com
016204.com

... Because you further devalue the extended names by mixing up the numbers to be more randomized. You are not just isolating the value based on length. You are also devaluing based on the randomization of the numbers in addition to the length extending outward. That's two variables instead of just one.

Click to expand...

And that is a good set to see the effects of a) memorability discount; b) meaning/luck discount

04, although has the meaning discount (04 is 4, just with meaningless discount) and luck discount for China/Chinese, so, let's say, instead of $5M, it would fetch $3M or so.
064 is similar in that sense and could get around 0.3M-0.5M
But should you have 640.com in your example, it would not get discount for meaningless, although 4 might hurt luck thing, being a "round" number would help. So it could easily sell for 0.6-0.8 range, above 064.com but below 989.com territory.

@21x20

21x20 said:

Part of what I'm getting at is that you can't easily compare apples to oranges and expect the same results.
IMHO, but this set is apples to oranges:

04.com
064.com
0164.com
04164.com
016204.com

... Because you further devalue the extended names by mixing up the numbers to be more randomized. You are not just isolating the value based on length. You are also devaluing based on the randomization of the numbers in addition to the length extending outward. That's two variables instead of just one.

Click to expand...

I respectfully disagree. I am not proposing to compare one list to another. I just think it's important to look at the top of the list, and then at the bottom of the list, and then average the price, to get to the more realistic results.

If you are basing your analysis only on 9s and 8s it's a waste of time as prices are extremely off balance, since most domains do not actually contain only 9s and 8s.... On the other hand - there are many more domains, which contain the random numbers I listed in my example set, and those are generally speaking what investors own and trade.

heavend said:

@21x20

I respectfully disagree. I am not proposing to compare one list to another. I just think it's important to look at the top of the list, and then at the bottom of the list, and then average the price, to get to the more realistic results.

If you are basing your analysis only on 9s and 8s it's a waste of time as prices are extremely off balance, since most domains do not actually contain only 9s and 8s.... On the other hand - there are many more domains, which contain the random numbers I listed in my example set, and those are generally speaking what investors own and trade.

Click to expand...

Agree to disagree then. In any thesis, the variables have to be considered. For example, you wouldn't try to compare tires using two different makes of car. You'd use the same car and put two different kinds of tires on it -- or you'd used the same tires on two different cars. But you would NOT put different tires on different cars and then try to come to any conclusion about the tires or the cars.

Likewise, in their own right, 989.com and 064.com are different beasts of quality within the NNN length genre. You can look at them alone and say quality matters in pricing within the NNN length.

Meanwhile... 989.com vs 016204.com are different tires on different cars. You can not discern anything separately about length or quality by looking at these two side by side.

They shouldn't be directly compared to come to any conclusions. And so similarly, modifying the top number to 064.com doesn't make for a reasonable comparison on length alone because iMHo the quality of 064.com itself is different from 016204.com. And since you're tossing quality into it, you're not properly examining the thesis being proposed.