TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Renaissance Technologies

299 pointsby pvsukale3over 5 years ago

35 comments

smabieover 5 years ago
A great book about RenTec and Jim Simons came out recently: The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution. Everyone in the industry worships Simons and RenTec as practically god-like. What they&#x27;ve managed to do shouldn&#x27;t really be possible and is out of this world. According to Wikipedia over a 20 year period between 1994 and 2014, RenTec realized an 71.9% annualized return in their internal Medallion fund. And before anyone says that&#x27;s a fluke, let&#x27;s calculate the probability of such a &quot;fluke&quot; using some simplifying assumptions:<p>1. S&amp;P 500 returns are log-normally distributed with a log-return of 10% and volatility of 10% 2. RenTec made a mean annualized return of 71.9% over 20 years 3. The returns of RenTec are independent from year to year<p>We get a P&lt;1.64E-14. Therefore, I think we all can agree that we can reject the null hypothesis of RenTec having no alpha.<p>And before someone says that it&#x27;s a scam, there&#x27;s no outside investors in the Medallion fund anymore, so if it is a scam, they would only be scamming their own employees. And if that was the case, I think we would know.
评论 #21506118 未加载
评论 #21506183 未加载
评论 #21507907 未加载
评论 #21506071 未加载
评论 #21506065 未加载
评论 #21512087 未加载
评论 #21507575 未加载
评论 #21508758 未加载
评论 #21511457 未加载
评论 #21510957 未加载
评论 #21508086 未加载
评论 #21506101 未加载
评论 #21506263 未加载
评论 #21510261 未加载
评论 #21506616 未加载
sbu_throwawayover 5 years ago
Here&#x27;s some inside baseball:<p>I recently graduated from Stony Brook University where Jim Simons chaired the math department in the 1960s. He left to start Renaissance Technologies which is located 1 mile down the street from campus. Their influence is everywhere. We have a Simons Center for Geometry and Physics ($150m building). I take classes in Frey Hall (Robert Frey used be managing director at Rentec). Our med school is literally the &quot;Renaissance School of Medicine.&quot; Our athletic facility is the Walter J. Hawrys Campus Recreation &amp; Wellness Center (named after Jim&#x27;s father in law). etc. etc.<p>Children of Renaissance employees come to campus on weekends to train for USAMO and play chess. I met Jim a couple of times at alumni events and what stood out is how humble and normal and down to earth he seemed. He is also a chain smoker and was lighting up indoors(!) but of course no one can say a word to him lol. He&#x27;s been making &gt;$1 billion a year for like the past 20 years. To put it in perspective: an average code monkey like me making $250k at a FAANG would have to work for FOUR THOUSAND YEARS to make how much he makes in a year. It&#x27;s unreal.
评论 #21508061 未加载
评论 #21511815 未加载
cjkenover 5 years ago
While I understand there are moral and ethical complications with the vast sums of wealth and influence attached to RenTec, the tone in these comments is disappointing...<p>There is no fraud at RenTec, and there is nothing magical about what they do. It&#x27;s simply an amazing technical and scientific organization, operating with almost unthinkable efficiency and scale.<p>I haven&#x27;t read the book, but I&#x27;m pretty sure this isn&#x27;t in it: I&#x27;ve been told by reliable sources that if you pick any one strategy (however you are able to separate&#x2F;define that) from the Medallion portfolio and it will not be individually remarkable--probably less than a 1.0 Sharpe (net of trading costs, gross of fees). There are countless hedge funds with equally performant models and sub-strategies.<p>What makes RenTec different is their ability to generate orthogonal strategies at a remarkable scale. They generate more ideas with less philosophical and empirical overlap than anyone else, by a wide margin. The software and theory required to do this for a fund as large as Medallion is absolutely as rare and valuable as their returns have proven.<p>Wrote more about it here: <a href="https:&#x2F;&#x2F;www.bridgealternatives.com&#x2F;medallion-isnt-magic-probably&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.bridgealternatives.com&#x2F;medallion-isnt-magic-prob...</a>
评论 #21509797 未加载
评论 #21509165 未加载
txtover 5 years ago
I use to service pools back in high school on the north shore of long island. And I remember doing Simons house, unbelievable property. The house keepers house was 5x bigger then mine..and the pool was massive, right on the edge of a cliff overlooking the long island sound. Sry if its offtopic but this read made me think of it!
评论 #21506921 未加载
评论 #21509807 未加载
gibyboover 5 years ago
Note that RenTec also runs two other funds that are larger than the Medallion Fund, but both under perform the index.<p>On a completely unrelated note, if I were interested in creating a fund that appeared to have market beating returns for decades and I wasn&#x27;t concerned about the legal consequences, here&#x27;s one way I might do it:<p>I would create fund A and B and seed them with some initial capital. For fund A, I would create a machine learning model that took in lots of opaque parameters and hire a team of very smart mathematicians and computer scientists to optimize this model to perform profitable trades. In order to easily generate alpha, I would subtlety feed in some parameters that were correlated to the future market moving actions of fund B. Since I also run fund B, I uniquely have this information.<p>I would use the performance of fund A to woo outside investors into investing their money into fund B. In order to prevent fund A from overtaking fund B in asset value and thus diminishing the value of the subtle signals, I would close the fund to the public when it got sufficiently large and periodically distribute its assets to its investors.<p>In order to avoid any of my employees from eventually figuring out what&#x27;s going on and reporting me, I&#x27;d incentivize them to avoid looking too closely by requiring that they invest a large part of their income into the fund with a long vesting date. I&#x27;d also require them to sign a long term non-compete so that they cannot work anywhere else in the financial industry if they leave.
评论 #21506368 未加载
评论 #21506481 未加载
评论 #21506347 未加载
评论 #21506540 未加载
评论 #21506964 未加载
评论 #21509256 未加载
评论 #21506693 未加载
评论 #21506466 未加载
评论 #21506395 未加载
评论 #21506380 未加载
mellingover 5 years ago
There was a book published last week on Jim Simons and Renaissance Technologies:<p><a href="https:&#x2F;&#x2F;www.amazon.com&#x2F;Man-Who-Solved-Market-Revolution&#x2F;dp&#x2F;073521798X&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.amazon.com&#x2F;Man-Who-Solved-Market-Revolution&#x2F;dp&#x2F;0...</a><p>The author was interviewed on the Masters in Businesss podcast:<p><a href="https:&#x2F;&#x2F;www.bloomberg.com&#x2F;news&#x2F;audio&#x2F;2019-10-30&#x2F;gregory-zuckerman-on-the-quant-revolution-podcast" rel="nofollow">https:&#x2F;&#x2F;www.bloomberg.com&#x2F;news&#x2F;audio&#x2F;2019-10-30&#x2F;gregory-zuck...</a><p>Simons is a serious mathematician:<p><a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Jim_Simons_(mathematician)" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Jim_Simons_(mathematician)</a><p>He won the Oswald Veblen Prize in 1976:<p><a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Oswald_Veblen_Prize_in_Geometry" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Oswald_Veblen_Prize_in_Geometr...</a>
评论 #21506013 未加载
deepnotderpover 5 years ago
Someone&#x27;s going to say this eventually, so it may as well be me.<p>Rentech is not the only hyper successful fund. There are others, like TGS management (<a href="https:&#x2F;&#x2F;www.google.com&#x2F;amp&#x2F;s&#x2F;www.cnbc.com&#x2F;amp&#x2F;2014&#x2F;05&#x2F;09&#x2F;mystery-13-billion-philanthropists-revealed.html" rel="nofollow">https:&#x2F;&#x2F;www.google.com&#x2F;amp&#x2F;s&#x2F;www.cnbc.com&#x2F;amp&#x2F;2014&#x2F;05&#x2F;09&#x2F;mys...</a>) that are just as successful and who you&#x27;ve never heard of.<p>What rentech has done is to have built an excellent data processing engine that automatically extracts signal from noise. Other, much more secretive, funds have done this too.
评论 #21506732 未加载
评论 #21511744 未加载
评论 #21506719 未加载
graycatover 5 years ago
Simons hired &quot;the best&quot; people?<p>On who Simons hired: IIRC there is an interview with Simons where he explains that he hired good mathematicians, etc. who had published interesting work or some such. In particular, he required no MBA or business degrees, experience with accounting, reading corporate annual reports, SEC filings, or business.<p>So, my guess would be that at least for the early years of Simons&#x27;s work, such people would not be very welcome in the rest of <i>Wall Street</i> and usually not considered the &quot;best&quot; or even good at all.<p>Point: That he hired &quot;the best&quot; has to do with what criteria one is using for <i>goodness</i>!<p>By the way, in this thread there is mention of Thorpe! So, yes, his <i>Beat the Market</i> had in the back that some argument was from &quot;measure theory&quot;. So, before my Ph.D., I was working carefully through Royden, <i>Real Analysis</i>! I was getting through the two exercises on upper and lower semi-continuity when I went to grad school. Right away I took the 700 level course in analysis (measure theory and functional analysis) and probability. The prof was a star student of E. Cinlar, long head of Operations Research and Financial Engineering at Princeton. Got to meet him once! So, that background should be of interest on Wall Street? After some good efforts, not as far as I could tell although then I didn&#x27;t yet know of Simons.<p>It is true that the year I got my Ph.D. near DC, the ORSA (Operations Research Society of America) conference was there and I submitted a resume. I was 110% busy taking care of my (later fatally) sick wife so didn&#x27;t go but some friends said that on some bulletin board I had lots of requests from Wall Street! Gee!<p>At one time I interviewed at Morgan Stanley and mentioned my paper on multi-variate, distribution-free anomaly detection. A computer systems management guy was a little interested. Another guy wanted me to give him an outline of the paper; I did and mentioned that I&#x27;d like to work on applied math of trading -- his reaction seemed to be that neither he nor Morgan Stanley nor anyone on Wall Street would be interested in any such thing!<p>Later a Bertsekas student from MIT gave me the rumor that Simons also liked hiring Russian mathematical physicists!<p>Point: It seems to me from a distance that Wall Street has been slow to hire people like Simons did.<p>On what Simons did: There&#x27;s an old comment that a good casserole is like a good marriage -- only the people who made it really know what is in it. My guess is that this remark applies also to Simons! In particular I doubt that we should regard him as a <i>quant</i> much like the rest of Wall Street.<p>Bluntly, I suspect that there&#x27;s no telling what the heck Simons did: E.g., there&#x27;s no law that he had to have much contact with the Markowitz work, the Sharpe work, the CAPM (capital asset pricing model), econometrics, classic time series analysis, the Thorpe work, the Black-Scholes work, the Martingale convergence theorem, the Doob decomposition, controlled Markov processes, stochastic differential equations, the Brownian motion solution to the Dirichlet problem -- he could have been doing things totally different!<p>Heck, quite generally it&#x27;s beyond super tough to be on the outside trying to know what&#x27;s going on inside: E.g., I got interested in cooking, once took my wife to André Soltner&#x27;s Lutèce, but never got a clue about how he did it! E.g., I made some progress with violin but never figured out how Heifetz did it! E.g., I made some progress in physics but never figured out how Einstein did it! E.g., I made some progress in math but really have no idea how Simons did it. It&#x27;s a Robert Frost one?? -- &quot;We dance round and round and suppose while the secret sits in the middle and knows.&quot;??<p>My approach to making money is not to try to get hired on Wall Street, out on Long Island, or near the commodity pits in Chicago or Kansas City but just to do my startup. There I don&#x27;t have to know all the ways people got rich in business -- it&#x27;s enough for me to find just ONE way!<p>I don&#x27;t have to know how all the cooking, music, physics, or math of the past was done -- it&#x27;s enough for me to do some math for the problem I&#x27;m trying to solve.<p>I don&#x27;t have to know every computer operating system, programming language, memory management or concurrency algorithm -- it&#x27;s enough for me to know enough computing to get the code running for my startup.<p>Since I&#x27;m able to be a sole, solo founder, I don&#x27;t have to convince people on Wall Street or Silicon Valley that I might make them money -- I just need to convince myself and then actually DO it!<p>And if I&#x27;m successful, my crucial core original applied math <i>secret sauce</i> should remain secret!<p>Point: I don&#x27;t believe I can figure out how Simons did it!
philshemover 5 years ago
I’m not a bot but I am always posting relevant and recent (2017) New Yorker articles that are worth reading:<p><a href="https:&#x2F;&#x2F;www.newyorker.com&#x2F;magazine&#x2F;2017&#x2F;12&#x2F;18&#x2F;jim-simons-the-numbers-king" rel="nofollow">https:&#x2F;&#x2F;www.newyorker.com&#x2F;magazine&#x2F;2017&#x2F;12&#x2F;18&#x2F;jim-simons-the...</a><p>And this interesting theme:<p>&gt; Foundations are not taxed, so much of the money that supports them is money that otherwise would have gone to the government. Scientific mega-donors answer to no one but themselves. Private institutes tend to have boards chosen by their founders, and are designed to further the founders’ wishes, even beyond their deaths. Rob Reich, a professor of political science at Stanford University and an expert on philanthropy, told me, “Private foundations are a plutocratic exercise of power that’s unaccountable, nontransparent, donor-directed, and generously tax-subsidized. This seems like a very peculiar institutional and organizational form to champion in a democratic society.”
评论 #21509879 未加载
superfishover 5 years ago
It blows my mind how much RenTech does with some ~300 employees. I recently had a phone screen with them (no offer otherwise I wouldn’t be writing this) and all of my communication was with this MIT math PhD. No HR, just the PhD.<p>Compare them with a Big N that has 10,000s of SWE’s. It’s a safe assumption that each engineer is individually less talented but even so, how do they iterate&#x2F;experiment with such few employees?
评论 #21511897 未加载
评论 #21509863 未加载
评论 #21511421 未加载
zozbot234over 5 years ago
&gt; <i>This article is about the American hedge fund management company. For technical advancements during the European Renaissance period, see</i> <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Renaissance_technology" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Renaissance_technology</a>
评论 #21506049 未加载
leto_iiover 5 years ago
I have to say, the first time I read about Renaissance I felt sadness at the thought that this much ability and such resources were put into essentially making money out of thin air, with very limited (if any) social benefits. I can&#x27;t escape the feeling that the world would be a better place if maths and physics phd&#x27;s would get to live a good life doing maths and physics, rather than being lured into quant finance.
评论 #21507359 未加载
analogkidover 5 years ago
The tone of many comments here is disappointing. I&#x27;m really surprised at the number of people suggesting illegal activity.<p>Why is it so hard to accept that someone did the math?
评论 #21507266 未加载
评论 #21508484 未加载
评论 #21507303 未加载
评论 #21511927 未加载
评论 #21511374 未加载
评论 #21507578 未加载
评论 #21509035 未加载
评论 #21511625 未加载
eternalny1over 5 years ago
Renaissance Explores Settlement as IRS Seeks Billions in Taxes<p><a href="https:&#x2F;&#x2F;www.bloomberg.com&#x2F;news&#x2F;articles&#x2F;2019-04-10&#x2F;renaissance-explores-settlement-as-irs-seeks-billions-in-taxes" rel="nofollow">https:&#x2F;&#x2F;www.bloomberg.com&#x2F;news&#x2F;articles&#x2F;2019-04-10&#x2F;renaissan...</a>
评论 #21506324 未加载
initiumover 5 years ago
This made me remember this interview with James Simons on Numberphile: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=sB_OdGGA450" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=sB_OdGGA450</a>
评论 #21506006 未加载
Tychoover 5 years ago
Who audits their returns? They have not had any outside investors for over a decade. I had not seen an update on their yearly performance for quite some time, and was surprised to see an updated table published in this book. So I&#x27;m curious, where did these numbers come from, and did RenTec confirm them, and did an independent party confirm them?
XnoiVeXover 5 years ago
So much negativity in this thread. We fear or doubt what we do not understand I guess.
评论 #21506432 未加载
rb808over 5 years ago
Do they still outperform? I can imagine 10 years ago they were ahead of everyone but now quant investing is everywhere I would be surprised if they have a big edge.
评论 #21507124 未加载
评论 #21506233 未加载
评论 #21506097 未加载
评论 #21506073 未加载
insiderthrowover 5 years ago
Either really good math or really good insider trading.
评论 #21506117 未加载
评论 #21506022 未加载
评论 #21505895 未加载
edwardover 5 years ago
Recently published book on this topic:<p>The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution by Gregory Zuckerman
anonuover 5 years ago
Nobody really knows how RenTech makes their money. There have been some interesting data points over time, mainly through lawsuits where interesting tidbits of info were divulged.<p>There are indications that the main rentech fund is a combination of high leverage and some potentially risky tax plays through structured products.
评论 #21508980 未加载
mason55over 5 years ago
A decent book about Jim Simons just came out last week<p><a href="https:&#x2F;&#x2F;www.amazon.com&#x2F;Man-Who-Solved-Market-Revolution-ebook&#x2F;dp&#x2F;B07P1NNTSD" rel="nofollow">https:&#x2F;&#x2F;www.amazon.com&#x2F;Man-Who-Solved-Market-Revolution-eboo...</a><p>I&#x27;m most of the way through it, it&#x27;s an interesting read
spyderover 5 years ago
A related promising company in this field is Numerai: <a href="https:&#x2F;&#x2F;numer.ai" rel="nofollow">https:&#x2F;&#x2F;numer.ai</a><p>Building a meta model from predictions submitted by the community (based on encrypted training data provided by Numerai).<p>More info in their video below. Also Howard Morgan the Co-Founder of Renaissance Technologies too is talking about them and a group of investors led by him invested $1.5m in the company:<p><a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=dhJnt0N497c" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=dhJnt0N497c</a><p>Their performance is not public, the only hint you can find about it is in their recent video about some backtests (which doesn&#x27;t mean it&#x27;s their real-world performance):<p><a href="https:&#x2F;&#x2F;youtu.be&#x2F;zeGx7gVgK0o?t=172" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;zeGx7gVgK0o?t=172</a>
评论 #21508167 未加载
naringasover 5 years ago
Do they produce&#x2F;create anything? or do they simply trade finacial assets in a way such that they accumulate profit?<p>and by &quot;simply trade&quot; I really mean &quot;trade in an incredibly sophisticated and clever way&quot;
评论 #21518748 未加载
account73466over 5 years ago
Performance of the 100bln fund is given at<p><a href="https:&#x2F;&#x2F;whalewisdom.com&#x2F;filer&#x2F;renaissance-technologies-llc" rel="nofollow">https:&#x2F;&#x2F;whalewisdom.com&#x2F;filer&#x2F;renaissance-technologies-llc</a>
ipsaover 5 years ago
So, assuming nothing against the law, how would they do it legitly? I am guessing:<p>- Treat the markets as a complex dynamical system and use the tools from statistical physics such as the Gibbs Ensemble, to derive internal states from input and output.<p>- Treat the markets as an encryption algorithm and use the tools from cryptanalysis, such as differential cryptanalysis: Even when unable to decipher the full algorithm (total break), one may still derive details and a subset of system functionality.<p>- They were probably the first to heavily use Hidden Markov Models (see Baum–Welch algorithm and the IBM speech recognition recruitment) and keep on the frontline with new machine learning algorithms (their deep learning revolution would have started 10-15 years before industry).<p>- They&#x27;d have an extremely solid backtesting pipeline, where any new feature can be stress-tested for signal. Features could be very arcane (% of mentions of the currency on neighboring state television) and are constantly (re-)added and removed: concept drift and market competition would gradually weaken signals, but fresh signals are added to keep the performance.<p>- All features are fed into a single final model (which may be an ensemble of many different forecasting techniques as to lower the variance). This model is very dynamic year-by-year (with just a few long-term signal features).<p>- Finally, I suspect there are strategies that only become available when you have 1 billion under control. In a physics sense: That is a lot of energy &#x2F; control theory experimentation budget. Normally, hedge funds would like to avoid feedback loops and their trades moving the markets, but I suspect there is a lot of money to be made when you can calculate in which direction the market would move when the system is deprived of - or infused with a jolt of energy. More hands-off: Buy for 1 billion in stock at market open, sell at market close. Buy signal will take a few hours to converge and result in a higher price, so you make a profit when you sell your portfolio to the very buyer&#x27;s market you created, causing a drop in price to complete the loop.<p>- The extreme returns for 2007&#x2F;2008 could be due to the increase in volatility of the crisis (you can make more money when there is a lot of action, and competitors suffer from human herd bias &#x2F; hysteria), but also, in part, due to them being the first to effectively exploit signals in growing social media platforms and search engines. A few years later it was public knowledge that gauging frequency and sentiment on Twitter was once a valuable signal.<p>- The NSA&#x2F;CIA type recruits would not work on industrial spying, but on cryptanalysis, (graph) data mining, OSINT, HUMINT, IMINT, and for the security of the firm (which probably runs a tighter security than the intelligence agencies of smaller countries).
评论 #21511134 未加载
评论 #21511090 未加载
gesmanover 5 years ago
Access to superior information under the disguise of math and science.
atlgatorover 5 years ago
I wish I was good enough to work for them. It must be fascinating.
kweinberover 5 years ago
You can’t separate Renaissance from Bob Mercer...the guy who helped fund Brexit, helped create Cambridge Analytica, the major founder of Breitbart news, and who got Trump elected mostly to avoid paying taxes.<p>He and his firm did more than anyone I know to make the world a more unstable place.
nfRfqX5nover 5 years ago
they talked about this on the Masters in Business podcast on Oct 30th. pretty interesting stuff.
评论 #21505846 未加载
trpcover 5 years ago
Jim Simons and David E. Shaw are legends who should have some HBO series about them. Both were researchers who left academia to beat the scumbags of wall streets in their own game with no finance background and they made unbelievably so much money in a very short of time that they would have been jailed or killed if they weren&#x27;t in the US.
评论 #21507438 未加载
viewbaseover 5 years ago
Back in late 2017, RenTec made an unusual move of raising capital to capitalize on &quot;market opportunities arising from Trump’s presidential victory&quot;:<p><a href="https:&#x2F;&#x2F;www.bloombergquint.com&#x2F;business&#x2F;renaissance-s-medallion-made-stunning-shift-after-trump-election" rel="nofollow">https:&#x2F;&#x2F;www.bloombergquint.com&#x2F;business&#x2F;renaissance-s-medall...</a><p>By Occam&#x27;s razor, it is more likely that Robert Mercer made use of his close ties with Trump administration to gain insider knowledge, than RenTec consistently having superior quantitative analysis than their competitors in an extremely competitive and saturated industry.
sjg007over 5 years ago
I dunno, I would take reliable weather reports a year out and try to predict crop prices. I am assuming demand will be linear. Maybe you can do this within the US, maybe not.
champagnepapiover 5 years ago
Here is a link to the book recently written about Jim Simons and RenTech <a href="https:&#x2F;&#x2F;www.goodreads.com&#x2F;book&#x2F;show&#x2F;43889703-the-man-who-solved-the-market" rel="nofollow">https:&#x2F;&#x2F;www.goodreads.com&#x2F;book&#x2F;show&#x2F;43889703-the-man-who-sol...</a>
unchockedover 5 years ago
A 66% annualized return over 30 years - from a completely opaque investment strategy. Extraordinary result for sure, how did he do it?<p>Run by Robert Mercer, the money man who associated with prominent money launderers.<p>Based on my priors, Occams Razor says Rentech is a laundromat, not a hedge fund.
评论 #21506034 未加载
评论 #21509016 未加载
评论 #21507031 未加载
评论 #21506037 未加载