In today’s economic climate, where governments, businesses and households are more conscious than ever of financial constraints, it is more important than ever that we direct spending to the most cost effective areas – interpreting statistics will be key in this.

An example from the news this week is this article from the Financial Times, which argues that the Bank of England’s economic forecasts, produced at great expense, have been practically worthless for many years:

Despite having hundreds of economists working in the Bank, and the most sophisticated suite of economic models in the UK, the monetary policy committee’s forecasts since 1997 have achieved no better outcome than if the committee had simply predicted the average level for inflation and growth over the 13-year period.

This raises the inevitable question: shouldn’t the Bank of England stop producing such forecasts, and, if it can’t justify employing so many economists on other grounds, stop wasting public money (i.e. yours and mine!) on this?

PS You can follow getstats on Twitter @RSSgetstats . I’m interested in using statistics to help decide where government cuts should go.

In response to a demand for explanations from members, the SOA issued a statement on its website on 22 July (SOA Board Addresses Questions Regarding Nominating Process).  Mr Bakos immediately challenged the accuracy of that statement, claiming that it was misleading and contained “spin” designed to give SOA members an incorrect impression (namely that the Board’s position was clearly supported by the Bylaws) whereas instead Mr Bakos’ view (and that of his supporters) is that the Board’s justification in refusing to add his name to the ballot is not in line with the Bylaws, but instead inline with a Board policy (including one introduced less than a year ago imposing a minimum of 2 and a maximum of 3 candidates for President-Elect “ordinarily”) which cannot override the Bylaws and seems (in the case of the most recent policy) not to have been ratified by SOA members.  Astonishingly, no logical rebuttal of Mr Bakos’s criticisms has yet been provided by the SOA leadership, despite repeated requests.

This begs the question: if the SOA leadership is confident that the logic of their position is sound, why have they not provided such a rebuttal, which would go a long way to defusing criticism and reassuring SOA members?

Something does not seem right here.  As actuaries, we are subject to stringent professional codes of conduct, and this matter cannot be left in the confused state it currently is.  We have to rely on logic and evidence to justify the positions we take, and it doesn’t seem to be in keeping with that for important questions to be ignored by the leadership of an actuarial body.  It is sad that this controversy has arisen, but as actuaries, I believe we have a professional duty under our codes of conduct (both the US and the UK ones) to clear it up swiftly.  More on this soon.

The new UK actuarial  body, The Institute and Faculty of Actuaries came into being today (1 Aug 2010, congratulations to it!) so it seems an appropriate time to ask its leaders to take a fresh look at environmental issues.  I have contacted what seem to be the relevant organisations and shall report again later this month.

Microsoft's ReadyBoost feature: is it of any practical use?

The ReadyBoost facility in Windows 7 (also in Windows Vista): is it actually worth using, or a bit of a gimmick?

I’ve been using it on and off with 4GB external USB drives for the last 3 years or so and it seems to make no noticeable difference whatever!

I will be interested to hear from anyone who has found it makes any noticeable difference to their computer’s performance.

My review of Office 2007 v Office 2003

Excel 2007 Pivot Table/Chart

Overall, I like Office 2007 and much prefer it to Office 2003.  This is mainly because Excel 2007 has a lot of improvements over Excel 2003, whereas I can’t really say that I’ve found much significant difference between the 2003 and 2007 versions of the other products (Word, Outlook, Access, Powerpoint and Publisher).  (I should add that I don’t use Powerpoint very often, so there may be significant improvements there that I haven’t yet had need to use).

So why do I much prefer Excel 2007 to Excel 2003?  Here are the main reasons:

• having the facility to use many more rows (over a million instead of about 65,000) and columns (over 16,000 instead of only 256) is very useful for larger data sets for which otherwise one would have to use Access
• Pivot Tables and Charts are much easier to create and use (see here for a tutorial)
• The charts produced look great and can be produced more quickly (and, although Excel 2003 also had this, it is easy to produce charts which update automatically as more data is added)
• the View Side by Side feature is very useful and has improved relative to Excel 2003
• Table and worksheet formatting is much easier to do

Excel 2007 Automatic Chart

Here are some things that are not quite as good in Excel 2007 as they were in Excel 2003:

• 3D graphs don’t work with more than about 3,000 data points (they cause Excel to freeze, see here for details)
• the ribbon user interface takes some getting used to.  After 3 years of using it, I find that about 2/3 of the time it is useful (because it makes me more aware of what functionality is available because the toolbar buttons depend on what I am doing), and about 1/3 of the time it is actually unhelpful (either because things that were easy to find before are hidden away under Excel options, or because it takes more mouse clicks in a few cases to do things than before).  However, the solution to the last bit is to add the item in question to the Quick Launch toolbar by dragging it there.

My verdict so far (after 7 months) on Windows 7 (as opposed to Vista)

I think it is definitely an improvement, but not as much as Microsoft claim:

• the extra Windows keyboard shortcuts for moving Windows around and tiling them (Windows key with left and right arrows, up and down arrows) are very useful to move windows from different applications around quickly and to get them sized easily to make maximum use of the available deskspace (including dual monitors) – no more having to drag the window edges using the mouse, which took a lot longer and was error prone! (You can find a near complete list of the new Windows 7 keyboard shortcuts here)
• the Windows Sidebar no longer takes up part of the right hand side of the screen, but instead can be brought into view quickly when needed by using Windows key + spacebar (which is a toggle)
• one simple but very useful improvement (which has been missing for years) is that Windows Explorer and the File Open/Save dialog boxes now includes a Recent Places shortcut (see here for a screenshot) which saves a lot of mouse clicks

Oherwise, I have to say that I have not found Windows 7 any faster to startup and shutdown than Windows Vista was, despite Microsoft’s claims about this.

Note that the Windows Photo Gallery, a good addition to Windows Vista, is no longer present in Windows 7.  However, that is not a problem because you can download Windows Live Photo Gallery free from Microsoft, which is actually an improved version of the Photo Gallery in Vista (so I recommend you download and use that even if you are still on Windows Vista or XP).  Why is Windows Live Photo Gallery better?  Because it has an improved user interface, with more options to Fix (edit) photos, including straightening/adding black and white effects, and the ability to create panoramic photos.

One disadvantage is that Windows Media Player 12, the version which comes with Windows 7, is not as good as Windows Media Player 11 in one important regard: it is much harder to edit the properties of tracks within an album because the Advanced Tag Editor is no longer there in version 12 of the Player!  Microsoft: what were you thinking when you removed that without supplying a good replacement for Advanced Tag Editor?

At last, a Recent Places link in Windows Explorer and Common Dialog boxes!

Windows Live Photo Gallery

I now look forward to the outcome of the vote, which we will know on Tuesday evening 25 May 2010.

I made the following post yesterday on the official merger forum (NB member log in required) in response to a post made by the Institute President, Nigel Masters:

(Post by Institute President on 20 May 2010)

“In the new governance documents Rules require a member vote and Regulations do not. We intended Regulations to cover all the things that the Council at present (given our desire not to shift the balance of power) can introduce without member consent. Although in our consultation on the governance documents 83% of respondents said they were happy with that approach, there have been some concerns raised that in future this could be used to introduce thing that members might feel they should have a say in. We would like to address that concern.
Going forward, as the leaders of the Profession, we would like to see more engagement between members and Council and for us to take advantage of new technology to increase that dialogue. We intend to introduce the means and the processes to enable members to do this easily, even though in our governance consultation only 45% of respondents thought the introduction of a consultation process was needed. We want to facilitate members in commenting on new Regulations before they are adopted and for members to have an easy process to follow if they have concerns. Joint Councils endorsed that principle at their meeting on 19 May 2010. As we were short of time we asked our lawyers to draft such a process which can be found on the Q&A section of the microsite and here. Currently it reads very “legalistically” and we will simplify the language and ensure it is in line with other processes such as those for introducing Standards – but it is good enough to give you the general idea. We intend that all such material will be easy to find in one place- the governance manual, also on the website.
Nigel Masters, Institute President
Also on behalf of Ronnie Bowie, Faculty President and Jane Curtis, Institute President-elect

(My post in response on 20 May 2010)

I think this is a very welcome development. I and some others have been concerned that the Regulations area reduced members’ rights by only allowing them to change Regulations they didn’t like by having both to successfully requisition a Special General Meeting and then get appropriate Resolutions passed at that SGM.  The changes announced today by the Presidents and the Institute President-Elect, in my view, go a long way to removing that objection (although Appendix 2 [proposed areas of exclusion for the new consultation process] probably merits further study with regard to some of its elements) by:

1. Allowing members to request that Regulations be put under the same consultation process as Rules, by 50 members making such a request (with what seem to be good electronic procedures for the necessary publicity to be made easy for any dissenting member).  Note: it is much easier for 50 members to agree to make such a request, than for 50 members to requisition an SGM – for the latter, members need not only to agree to requisition a meeting, but also the agree on the precise wording of any Resolutions.  Member requisitioned SGMs have been very rare, and one reason for this has I think been the difficulty of getting even 10 members to agree on the precise wording of Resolutions

2. Re-establishing the principle (and providing an efficient mechanism fit for the electronic age [via prompt announcements on the Profession's website, in general e-Alerts to members, and in The Actuary magazine]) that members are to be consulted regularly (as opposed to ad hoc, for specific big projects such as the merger) on important matters and can add short comments (which can include links to websites controlled by the members with longer comments).

I attended one of the Professionalism Courses today and was pleased to hear that the Institute & Faculty’s new website in July will allow members to add comments in many more areas.  These are the sort of steps towards increased openness and transparency that I have been hoping for.

I’ve had a long day, so more tomorrow on this.

Today, I made the following post on the official forum:

I stated after the failed July 2009 merger vote that I would support a revised merger with a better name and which did not significantly reduce members’ rights.

After reflection, and given that Councils have effectively (with yesterday’s statement) removed my remaining significant objection to the proposed merger, I now support the 2010 merger proposal. I have written to fellow FIDELIS members explaining this and informing them, consistent with my support for the merger proposal, of my resignation from FIDELIS.

The initial analysis is that the governance documents still include a significant transfer of power from members to the new Council, and that David Wilkie/FIDELIS’s Unification plan is simpler and achieves the same benefits whilst avoiding mutilating the Institute’s name and needlessly destroying the Faculty. It also maintains the balance of power between members and Councils.

Please do not vote until after you have had a chance to read our response which will be issued on the FIDELIS website (with link from here) in the next week or so.

The answer is at

http://forums.microsoft.com/MSDN/ShowPost.aspx?PostID=317267&SiteID=1

Hope this helps – it removed the error 430 “this class does not support automation” which seems to be given (very uninformatively) by VBA if the .net security doesn’t allow the dll to run.

(For ease of reference, and in case – as happens sometimes on the internet – either of the Google Answers or MSDN forum pages get removed, I am posting the question and my answer here in full below:)

Question (asked in March 2006):

We currently have a product which happily runs across a network by having various COM dlls installed on a shared network folder and various client pcs use these dlls via an VB6 exe and also Excel
(2000/XP/2003) VBA. We don’t have to install anything on the clientpcs apart from having to register the COM dlls (via regsvr32 “//server/sharedfolder/COMDll.dll”) so that Excel (which runs locally
on each client pc) knows where to find them (i.e. on the shared network folder).

We now want to start using .net dlls (exposed for COM via the “Make assembly COM visible” checkbox) for any new code. My question is, after installing the .net dlls on the server in the shared network
folder (//server/sharedfolder/), how can we most easily make them available to client pcs?

My understanding is that we need registry entries to be added on each client pc for each ClassID, ProgIDs etc exposed by the .net COM dll, the assembly location etc. This is one of the things that a setup
program would do if we installed our application locally on each client pc. But the whole point of the network installation is that we don’t want to have to install the dlls locally on each pc.

I’ve already tried using regasm “//serverpath/sharedfolder/DotNetCOM.dll” /codebase instead of regsvr32 to register the .net COM dlls on the client pcs. In fact what I have done more precisely is to use regasm
“//serverpath/sharedfolder/DotNetCOM.dll” /codebase /regfile:”file.reg” to create a .reg file (file.reg) which can be run on each client pc. This is probably better because otherwise problems
will arise if the version of regasm on the client pcs is different from that on the server.

However: after testing this across the network using Excel 2003 VBA, I get an error message “Error 430: class does not support Automation or does not support expected interface” the first time I
instantiate an object in VBA:

Dim obj as MyClass

set obj = New MyClass ‘ error 430 occurs here when the dll is across network (but – see below – code works fine if dll is local)

Interestingly the error 430 disappears and VBA interacts happily with the dll (I can call its methods etc.) if DotNetCOM.dll is moved to be local to that pc (e.g. C:\localfolder\DotNetCOM.dll) and regasm rerun
(or the file.reg altered so that the path is changed to the local one, or simplest of all, simply changing the codebase key value in the registry to point to C:\localfolder\DotNetCOM.dll rather than
//serverpath/sharedfolder/DotNetCOM.dll).

Further information: I posted this question 3 days ago on an MSDN forum (see http://forums.microsoft.com/MSDN/ShowPost.aspx?PostID=317267&SiteID=1), but so far it doesn’t look as if I’m going to get an answer from
there, and this is an important issue to us, hence am posting it here. At the moment the DotNetCOM.dll assembly is not signed, and has not been put into the GAC (Global Assembly Cache). This is because this
is not necessary as shown by the fact that calling DotNetCOM.dll from VBA works fine with DotNetCOM.dll manually installed to local folder C:\localfolder\DotNetCOM.dll as shown above. DotNetCOM.dll is a
Visual Studio 2005 C# dll which I created to test this (working across the network) issue, with a very simple interface, with one method, Add which adds two doubles together. I also know that the client machine
I am testing on has the .net 2 framework runtime expected by DotNetCOM.dll, mscoree.dll version 2.0.50727. (I realise that if the client pc only had an earlier version of the .net framework, then as
well as creating the registry entries via file.reg, I would also need to install the newer version of the .net framework, but that is obviously not contributing to the problem here since a) the .net
version on the client pc is up to date, and b) the dll works fine when placed locally). The fact that the dll works locally presumably means that the fact that so far I am not using a shim, i.e. the
InProcServer32 default key value is mscoree.dll, is not an issue either. Another point is that marshalling, multi threading, cross processes should not be an issue here, since the dll is an extremely
simple one (it is not an exe), and should simply be running inside Excel.exe’s process. I had always thought that to Windows, a shared network folder is simply another drive, so I can’t understand at the
moment why my simple dll works locally, but not when moved to the shared network folder?

Answer:

(Update: I eventually solved my problem, and posting it here because lots of people seem to have the same problem and ask me whether I found a solution).

In fact, contrary to what the documentation says (and any warning messages you may get from caspol), it is NOT necessary (with .net 1.1 or 2.0) that the dll be strongly named, all that matters is that the dll has the necessary permissions to run.

The problem in my case turned out to be .net security. I had to grant security policy on the pc on which I wanted to run the software to grant full trust to the URL of the folder where the dll resides file://server/shareddlls/* . I used caspol to do this. (Granting full trust to assemblies in that location signed with the appropriate strong name will also work.)  (NB you need to be careful to make sure that you add the permission in the correct zone, e.g. intranet would normally be the zone for a shared server, but by default, I found that Internet Explorer 7 was treating //server as part of the internet zone rather than intranet, until I changed a few settings.  You can find out which zone your pc is treating //server by navigating to it in Windows Explorer and looking in the bottom right of the status bar, it should say intranet or internet).

Hope this helps,

Patrick.

For ease of reference, I also put a copy of the two relevant texts (so far) below

(David Wilkie’s letter in March to The Actuary)

A simpler plan

It is sad that Councils are repeating last year’s merger mistake, proposing too much at once. The only changes since last year’s proposals are the name, and bits of the draft governance document. Institute Council claims that unification requires a new Charter. Faculty Council claims that unification requires the Faculty to be wound up. Neither claim is true.

A much simpler unification plan is for the Institute, by a special Bye-law or otherwise, to take in all Faculty members in their corresponding grades, and arrange for some Scots to be on Council. And then for Faculty Council to relinquish its professional body activities to the Institute, but keep the Faculty in existence as a Scottish actuarial society, which would run Sessional Meetings, and might do more. Subscriptions would need to be adjusted appropriately.

Faculty members that did not like being in the Institute could resign, and should be allowed to stay in the Faculty; they might not be able to get practising certificates, but could still practise as actuaries (as any of us could without being in either body). Faculty members that were content to be only in the Institute could resign from the Faculty.

This is simple, and would need a two thirds majority in an Institute vote, and no formal vote in the Faculty, but to get members’ agreement in principle would be very advisable.

It would not be a takeover of the Faculty, but a return home for the successors of those who left the Institute in 1856.

The Institute could change its Bye-Laws and the Faculty its Rules at leisure. Neither Charter requires any change.

Members should be given the opportunity to say whether they prefer this simpler plan.

To read more and to comment on this plan go to http://adwsplan.wordpress.com.

David Wilkie, FFA, FIA

(statement issued by Ronnie Sloan FFA on 12 Mar 2010 on behalf of a group of FIDELIS members as a proposed alternative to the merger plan being put forward by the Councils of the Institute and Faculty of Actuaries)

Unification : better than Merger
Although the overall merger vote last June failed, Faculty members in fact voted in favour, and we respect their wishes. However, we believe that these wishes would be better served not by an outright merger, but by a form of unification that would retain a continuing Faculty.

Our unification proposal also offers considerable advantages to Institute members.

We therefore invite all members to consider the merits of our case, which is as follows:-

1. The Core Issue

The merger proposes the needless destruction of the 154-year old Faculty of Actuaries in Scotland and its replacement by a sub-committee of a London-based body. Paradoxically, it is contended
that this will somehow strengthen the role of the actuarial profession in Scotland!

On the contrary, a properly managed, albeit trimmed down (see later) continuing Faculty could perform similar functions to the proposed Scottish Board, but would thereby enjoy higher status
within Scotland, and hence have greater influence in Scottish financial affairs. Moreover, it could then easily be fully reactivated in the event of greater devolved powers or Scottish independence.

As regards the 162-year old Institute of Actuaries, the merger proposes the wanton mutilation of its name into a compromise whose initials are inconsistent with members’ continued use of FIA.
So, such degree of disruption hardly seems justified for a modest 17% increase in its membership.

2. Pro-Merger Arguments advanced by Councils

These are largely assertions without evidence or analysis, of which the two principal ones are :-

• The existence of two professional UK actuarial bodies is confusing. We need to simplify matters, and speak with one voice.

The general public and the media care not a jot whether an actuary is an FFA or an FIA. And the authorities – do they not understand our separate embodiment? Of course they do!

And the proposed ‘simplification’? – to call the new body ‘the Institute and Faculty of Actuaries’, but with members using either FFA (relating to a body that no longer exists) or FIA (same body, but a different name). If the press and the public were confused before, can this really help?

As to ‘one voice’ on matters of public interest, we already have a mechanism – when we agree – for doing so on behalf of both bodies. But what if we didn’t always agree, or needed to address purely Scottish issues? Merger removes that flexibility on the UK domestic front. And why cut back unnecessarily our UK votes on the wider international actuarial stage through the IAA?

• The present cumbersome relationship between the two bodies hampers efficiency.

Unquestionably true – but entirely the fault of both Councils for allowing their joint w orking to depart from the originally envisaged model, and then for the last few years obsessively pursuing merger instead of tackling the problems they had themselves allowed to develop!

However, we have a much simpler suggestion than the proposed merger – see below
Moreover, we believe our suggested approach to be not only simpler, but better, and which we call:-

3. The Fidelis Unification Plan

Institute Council still seem to claim that merger requires a new Charter; and Faculty Council that merger requires the Faculty to be wound up. Neither claim is true, as can be demonstrated by considering our simpler, and we think better, Unification Plan, the key points of which are:-

3.1 The Institute would not need to change its name, Charter or governance, and need merely take in all Faculty members in their corresponding grades, and also arrange for some FFA members to be on its Council (perhaps similar to the Scottish Board).

3.2 Faculty Council would relinquish its ‘professional body’ activities in favour of the Institute, but continue to run Sessional Meetings as at present, which would be published in BAJ; and otherwise act as proposed for the ‘Scottish Board’, including the one-off endowment, but with far greater authority and professional respect within Scotland.

3.3 Supervision of ‘professional body’ activities of all Faculty members would be transferred to the Institute by some such simple mechanism as a special Bye-law.

3.4 Asset and liability transfers would be implemented as envisaged for merger, with minimal or no specific staff required for the Faculty in its capacity as a Scottish actuarial society.

3.5 Faculty Council would be smaller and there would be no need for Joint Council meetings; and most of the current joint committees could continue as Institute-only committees, but of course with former or current FFA members included as appropriate.

3.6 Provision would be made for FIAs in Scotland also to join the Faculty and to participate fully in its Scottish activities, for which no significant (or any) extra subscription would be asked.

3.7 And vitally, the Faculty would retain its existing very flexible Royal Charter and the right to reactivate its full ‘professional body’ activities in the event of much greater powers being devolved to a Scottish Government, or in the event of full Scottish independence.

Further details of this unification proposal will be provided at a later stage.

4. Conclusion

We believe that such a continuing Faculty would provide the same valuable contribution to actuarial and commercial life and thinking in Scotland as in the past, which would be far better than for actuaries in Scotland to be dependent on the proposed Scottish Board within the I&FoA.
This proposal should also appeal to Institute members, who would avoid having their historic body’s name needlessly changed and their FIA designation rendered largely meaningless.

Finally, we believe that Faculty members, especially those in Scotland, should be given the opportunity to say whether they would prefer a continuing Faculty along these lines rather than the irreversible abolition of their historic professional body.

R K Sloan FFA 12 March 2010
(spokesman on behalf of FIDELIS)
email: ronnie at sloan42.fsnet.co.uk www.fidelisdefence.com

The above statement of 12 March 2010 has been specifically endorsed by the undernoted Fidelis members:-

David Wilkie FIA FFA (past V-P of both Institute and Faculty)
John McCutcheon FFA (Faculty President 1992-94)
Malcolm Murray FFA (Faculty President 1994-96)
Paul Grace FFA (Faculty President 1996-98)
Fraser Low FFA AIA (Faculty President 1998-00)
David Purchase FIA (past V-P and Hon Secy of the Institute)
Gordon Bannerman FFA AIA
Patrick Lee FIA
Mike Lunan FFA
Brendan McBride FFA, Inst Affil
Jon Spain FIA
Ronnie Sloan FFA

Our spokesmen can be contacted as follows:-
Until 31 March 2010:- Ronnie Sloan ronnie at sloan42.fsnet.co.uk
and from 1 May
During April 2010:- David Wilkie david.wilkie at inqa.com