tag:blogger.com,1999:blog-34323687.post9182113667691732677..comments2024-03-19T03:05:57.184+00:00Comments on Macro Man: 20 QuestionsMacro Manhttp://www.blogger.com/profile/12324967552369915949noreply@blogger.comBlogger41125tag:blogger.com,1999:blog-34323687.post-30589958916728487762009-06-20T07:02:43.745+01:002009-06-20T07:02:43.745+01:00No. 20 - Because it has been statistically proven ...No. 20 - Because it has been statistically proven that a cyclist is more likely to die stopping at a red light than riding through it.Andreahttp://www.velorution.biznoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-75743879890834379672009-06-19T16:36:45.557+01:002009-06-19T16:36:45.557+01:00Ok, let me put it this way. For my style of tradi...Ok, let me put it this way. For my style of trading, I would say Kelly is too aggressive and not even close to being a guidepost to anything. I certainly did not mean to suggest that you were a twenty year old with a maths degree and meant no offence at all. You don't have to be a twenty year old with a math degree to make unfounded assumptions about how markets play out in practice. Just think of all the short vol hedge funds that get a regular hosing and the funds containing Nobel prize winners (LTCM) which got caught in compromising positions. Anyway, no offence was meant and I hope none taken, just giving my opinion.thetradernoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-48689788561051738592009-06-19T15:43:08.110+01:002009-06-19T15:43:08.110+01:00I think we're saying exactly the same thing. ...I think we're saying exactly the same thing. My confusion stems, I posit, from the fact that I said it first ("Kelly is limited, since it presumes equally perfect certainty in all estimates... unexpected correlation changes will blow you up better than anything I know"). Perhaps you saw the phrase "Kelly criterion," assumed I was a twenty-year-old with a math degree, and stopped reading.<br /><br />Fuller disclosure: my never-blown-up track record starts before the 1997 Asian currency crisis. However, I do have a math degree.wcwhttps://www.blogger.com/profile/16307608293310560164noreply@blogger.comtag:blogger.com,1999:blog-34323687.post-46077537448361750732009-06-19T09:32:44.606+01:002009-06-19T09:32:44.606+01:00wcw,
As a trader, the problem I have with Kelly i...wcw,<br /><br />As a trader, the problem I have with Kelly is you can never really know with absolute certainty whether your next trade will be a win or a loss. Nor can you know if you may have a string of losses in a row. Kelly, applied correctly, could take you out of the game before your positive expectation has time to work itself out. That is the problem I have with Kelly in real-world implementation. Also, you cannot guarantee size of payoff either. These mathematical risk management criterion look great on paper but are difficult to apply in the face of uncertainty. It is a similar problem to the correlation one. Correlations work great until there is an exogenous event and everything moves together. Better to assume the worst and position size accordingly. Anyone who has traded for an extended period of time will tell you something similar, I would guess.thetradernoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-86734222179767178132009-06-18T19:47:42.690+01:002009-06-18T19:47:42.690+01:00If a real-world implementation or your back-of-the...If a real-world implementation or your back-of-the-envelope intuition of deservedly seminal game-theory-class Kelly is too aggressive, you might be too confident in your win-probability or payoff estimates. Probably both.wcwhttps://www.blogger.com/profile/16307608293310560164noreply@blogger.comtag:blogger.com,1999:blog-34323687.post-2880708669122191042009-06-18T18:30:42.687+01:002009-06-18T18:30:42.687+01:00Typo, meant Jun10 on rate inc.
Good point Greg, w...Typo, meant Jun10 on rate inc.<br /><br />Good point Greg, would have thought that a more streamlined regulatory system would have been top of the agenda in trying to prevent a build of systemic issues, instead we get more silos.<br /><br />Interesting points and moves today. But I always worry when the market gets overly focused on a single month's data.<br /><br />TGIFish, JLAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-78266162989458972082009-06-18T17:36:48.265+01:002009-06-18T17:36:48.265+01:00Question #21: Is a guy swatting a fly really news...Question #21: Is a guy swatting a fly really news? Is a guy taking his wife out to dinner really news?<br /><br />Good leaders are not surrounded by armies of yes-men. The news media in the States and Europe look at Obama like giddy school girls drooling over the boy band du jour.<br /><br />With no scepticism and no one playing devils advocate to his ideas -- Obama is destined to make terrible decisionsGregnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-22281390730250846002009-06-18T17:24:46.530+01:002009-06-18T17:24:46.530+01:00macroman --
chinese banks entered into the crisis...macroman --<br /><br />chinese banks entered into the crisis with a very low loan to deposit ratio, so they don't to conjure up funds out of thin air -- only to lend out deposits they already have.<br /><br />of course, the net effect is a very strong monetary stimulus ...<br /><br />china -- b/c of a liquid state banking system that is subject to administrative guidance -- has some policy tools other countries don't have.<br /><br />bsetser <br /><br />bsetserAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-21509482780693003572009-06-18T17:12:40.297+01:002009-06-18T17:12:40.297+01:00Kelly? that'll kill you in the real world, way...Kelly? that'll kill you in the real world, way too aggressiveAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-45352868864586936952009-06-18T17:07:55.198+01:002009-06-18T17:07:55.198+01:00I know nobody asked me, but in re: position sizing...I know nobody asked me, but in re: position sizing, I am a fan of always keeping the Kelly Criterion in the back of your head. Kelly is limited, since it presumes equally perfect certainty in all estimates, but it's the best guidepost I know.<br /><br />Also, I'd like to echo MM's note on changing correlations. Whether you're programming a black box or, like we, constructing portfolios by hand, unexpected correlation changes will blow you up better than anything I know.<br /><br />FD: I have never blown up, but every 10+% drawdown sure feels like a blowup.wcwhttps://www.blogger.com/profile/16307608293310560164noreply@blogger.comtag:blogger.com,1999:blog-34323687.post-64543989545691763672009-06-18T16:59:59.622+01:002009-06-18T16:59:59.622+01:00really loving this 10yr reversal
mpmreally loving this 10yr reversal<br /><br />mpmAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-26442209560075643062009-06-18T16:06:43.732+01:002009-06-18T16:06:43.732+01:00Yeah, we have daily VaR limits (ex-ante) and also ...Yeah, we have daily VaR limits (ex-ante) and also drawdown limits (ex-post.)<br /><br />However, I have also managed money in the same why in places that just hav nominal position limits and desired risk targets, but nothing more binding than that.Macro Manhttps://www.blogger.com/profile/12324967552369915949noreply@blogger.comtag:blogger.com,1999:blog-34323687.post-73553999779885222592009-06-18T16:04:25.699+01:002009-06-18T16:04:25.699+01:00would you mind if i asked do you have a volatility...would you mind if i asked do you have a volatility limit or target at least where you work?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-59047892336467732332009-06-18T15:58:24.799+01:002009-06-18T15:58:24.799+01:00Anon @ 3.36, I really have no idea in your example...Anon @ 3.36, I really have no idea in your example, because I don't manage money in that way (i.e., having only one trade.) I know from experience that averaging certain level of portfolio VaR usually translates into a certain level of portfolio level volatility given the way I construct portfolios (in which the sum of the individual position VaRs can be say 3 or 4 times the portfolio level VaR.)<br /><br />I can count on one hand the number of times I have had a limit-sized position in any one trade over the past five years, because for me to do so requires a confluence of factors that is extremely rare.<br /><br />As for correlation risk, that has been the source of most of my worst days in managing money...I construct a portfolio of un- or negatively-correlated trades, but then the correlations change and my ex-post risk ends up much higher than my ex-ante.<br /><br />I find that there are usually signs that this is in the offing, so I just try and keep a close eye on my book to see if realized starts ticking up without me adding new risk.Macro Manhttps://www.blogger.com/profile/12324967552369915949noreply@blogger.comtag:blogger.com,1999:blog-34323687.post-56846733234523984402009-06-18T15:58:02.180+01:002009-06-18T15:58:02.180+01:001) SPX 1000
2) No
3) As big as expected
4) No
5) V...1) SPX 1000<br />2) No<br />3) As big as expected<br />4) No<br />5) Very<br />6) Not soon enough<br />7) They are in league with the Devil<br />8) 100<br />9) When the Somali Pirates get of prison ;))<br />10) No<br />11) ECB<br />12) 2010<br />13) I wasn't aware there are any :)<br />14) Just macro man<br />15) 2010<br />16) No<br />17) No<br />18) Nowhere near toppy in gold, just positioning cleanout<br />19) yes, very soon<br />20) because they are immortal, didn't you get the memo?? :)<br /><br />-IvanovskiAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-87315739010822286642009-06-18T15:52:28.809+01:002009-06-18T15:52:28.809+01:00Again, very interesting. I guess it can be great ...Again, very interesting. I guess it can be great when you are right and you have your max position on, but quite destabilizing when you are wrong and you have your max position on. I suppose the scaling in and out takes care of that, in some ways. I wonder if an equal position size method would be any use - kind of hurts the ego a little bit though! But it does stop one from being too emotionally invested in a particular trade and acknowledges the possibility of being wrong.thetradernoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-26140604023844091632009-06-18T15:37:19.737+01:002009-06-18T15:37:19.737+01:00It really depends on the market environment, stren...It really depends on the market environment, strength of conviction, what my exisiting portfolio looks like, etc. I don't use hard and fast rules...sizing the trade for me is more art than science.Macro Manhttps://www.blogger.com/profile/12324967552369915949noreply@blogger.comtag:blogger.com,1999:blog-34323687.post-18580650691515991082009-06-18T15:36:34.544+01:002009-06-18T15:36:34.544+01:00I second that. Thanks for insight. At a reasonable...I second that. Thanks for insight. At a reasonable risk fund with 15% returns expected and 12% vol or so what do you think is the max percentage you should risk on a full sized trade. assume just 1 trade in portfolio at time to avoid issue of correlation which is a second question if you care to answer (how do you factor in correclation in total portfolio risk?)riskAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-2655699876127663152009-06-18T15:33:20.312+01:002009-06-18T15:33:20.312+01:00MM,
very interesting insight into your position m...MM,<br /><br />very interesting insight into your position management. How do you size each position in terms of percent of portfolio risked per trade or correlated trade?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-78617218968508397552009-06-18T15:13:06.530+01:002009-06-18T15:13:06.530+01:00Anon @ 1.55, what you ask for could either be a ve...Anon @ 1.55, what you ask for could either be a very long or a very short answer. The former would be the subject of a post in and of itself, so I'll opt for the latter.<br /><br />Essentially, I try to run my book as a portfolio rather than a series of trades. At any point I will have a given number of themes that I'll be following, and will scale into positions as they gain traction and scale out as they mature. I also try and use options where possible to mitigate the peaks and valleys of short term price action.<br /><br />I like to use technicals to initiate a position (i.e., where there is a point in reasonably close proximity where I feel the market is telling me I am wrong) and/or to get out. I also do a few tactical technical trades with modest results.<br /><br />I also always like to have cheap hedges in place that will perform disproportinately well if my underlying themes are proven wrong.<br /><br />Basically, I view portfolio construction as half art and half science. I think it's important to know the limits of contructs like VaR and correlation matrices; as it is, I am generally exposed to the risks posed by changing correlations.Macro Manhttps://www.blogger.com/profile/12324967552369915949noreply@blogger.comtag:blogger.com,1999:blog-34323687.post-21884297939265601572009-06-18T15:13:00.715+01:002009-06-18T15:13:00.715+01:00Would like to add question 21. Is it safe to sell ...Would like to add question 21. Is it safe to sell fx vol over the summer? The DNT calculators have been dusted off and my bookies are sending a plethora of 2-3m ideas. <br /><br />DCAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-57123867145312712892009-06-18T14:47:10.756+01:002009-06-18T14:47:10.756+01:001)SPX 800
2)Unlikely, but backup in mortgage rates...1)SPX 800<br />2)Unlikely, but backup in mortgage rates should warrant something.<br />4)Yes, see us hovering here then to 1.35<br />6)There needs to be a lot more questioning of China bulltards going on!<br />7)Not even sure they know. Perhaps they have a prop book?<br />8)Tricky but would say 50.<br />11)Close call BoE/ECB, think latter due to inflationary obsessions, not greeeen shoots.<br />12)Jun11?<br />13)When its safe to do LIBOR carry.<br />15)Personal saving rate 10% peak and poss permanent shift<br />16)ARMs, like CRE and jumbo issues, are lurking<br />18)Top in retail gold but not in the market price / USD deval / inflation hedge story. Still...<br /><br />Oh, Anon & Evaluation 30-31, please let up, one request is enough. <br />Would like to hear more from the bulls. mpm any care?<br />Cheers, JLAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-50818102531129805472009-06-18T13:55:11.159+01:002009-06-18T13:55:11.159+01:00effort side is understandable. But I do find the g...effort side is understandable. But I do find the great traders are better at selection and timing than the good ones so I would like to know your thought process behind tardes if you would be so good. do you have any definite rules that have to exist before you enter a directional rates or fx position? ie ... like always trade with a trend (already moved 5% or what ever that way). Second. what do you think is the best way to evaluate if you are getting better at selecting trades?...its not always just prrofits....do you prefer to look at any particular statistic about your historical trades to see if you are improving?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-34323687.post-71698270448908982062009-06-18T13:51:10.928+01:002009-06-18T13:51:10.928+01:00Anon @ 13.24, that copper trade was a punt based o...Anon @ 13.24, that copper trade was a punt based on a) the market's green shoot mania at the time of trade, and b)the technical set up. It only lasted about three days and is now long forgotten...Macro Manhttps://www.blogger.com/profile/12324967552369915949noreply@blogger.comtag:blogger.com,1999:blog-34323687.post-80754822105764877602009-06-18T13:48:43.849+01:002009-06-18T13:48:43.849+01:00Nemo, re: China- the local government borrowing sp...Nemo, re: China- the local government borrowing spree (and forthcoming threat of bankrputcy) is exactly what I was referring to. It's remarkable how the market is obsessed with fiscal eficits in the US and UK, and yet happily assumes that China can conjure Y4 trillion out of thin air....<br /><br />Evaluation: per the FAQs, the primary reason that I don't post positions is that it would require quite a bit more work than it used to because of the improved systems at my current shop. Having my brokers know my positions is a secondary consideration, but really, it's about not wanting to do extra work.<br /><br />There is still plenty of opportunity to discuss the implementation of given themes here in the comments section.Macro Manhttps://www.blogger.com/profile/12324967552369915949noreply@blogger.com